Prévia do material em texto
www.thelancet.com/haematology Vol 9 August 2022 e615 Review Lancet Haematol 2022; 9: e615–26 *Contributed equally Department of Pediatrics and Adolescent Medicine, Comprehensive Center for Pediatrics (L Bellach Cand Med, A Hassanein Cand Med, Prof A Repa MD, Prof K Klebermass-Schrehof MD, L Wisgrill MD, V Giordano PhD, Prof A Berger MD), Center for Medical Biochemistry, Max Perutz Labs (M Eigenschink Cand Med, D Savran Cand Med, Prof U Salzer PhD, Prof E W Müllner PhD), Medical University of Vienna, Vienna, Austria Correspondence to: Prof Angelika Berger, Department of Pediatrics and Adolescent Medicine, Comprehensive Center for Pediatrics, Medical University of Vienna, Vienna 1090, Austria angelika.berger@meduniwien. ac.at Packed red blood cell transfusion in preterm infants Luise Bellach*, Michael Eigenschink*, Abtin Hassanein, Danylo Savran, Ulrich Salzer, Ernst W Müllner, Andreas Repa, Katrin Klebermass-Schrehof, Lukas Wisgrill, Vito Giordano, Angelika Berger Premature infants commonly receive adult packed red blood cells (pRBCs) during their hospital stay. As adult erythrocytes differ substantially from those of preterm infants, transfusion of adult pRBCs into preterm infants can be considered inappropriate for the physiology of a preterm infant. An absence of standardisation of transfusion protocols makes it difficult to compare and interpret pertinent clinical data, as reflected by unclear associations between pRBC transfusion and complications related to prematurity, such as bronchopulmonary dysplasia, neurodevelopmental impairment, retinopathy of prematurity, or necrotising enterocolitis. The difficulty in interpreting clinical data is further increased by differences in study designs that either overestimate pRBC-associated complications of prematurity or have not yet been designed to directly link pRBC transfusions to their respective complications. Thus, neonatal transfusion practice has become an ongoing difficulty, in which differences in transfusion guidelines hinder the ability to generate comparable clinical data, and heterogeneity in clinical data prevents the implementation of standardised transfusion protocols. To overcome these issues, novel approaches with biochemical-clinical translational designs could enable clinicians to gather causal evidence instead of circumstantial correlation. See Online for appendix Introduction According to WHO, about 15 million babies are born prematurely every year, with a ratio varying between 5% and 18%, depending on the country of origin.1 Considering the amount of physiological, haemodynamic, and respiratory immaturity, prematurity is an acute and potentially life-threatening condition that implies specific complications of multiple organ systems. Broncho- pulmonary dysplasia,2 intraventricular haemor rhage, post-haemorrhagic hydrocephalus, periventricular leuko- malacia,3 retinopathy of prematurity,4 necrotising entero- colitis,5 and hypoxaemia and anaemia of pre maturity6 are only a few examples of such complications. Although advancements in neonatal intensive care have reduced mortality to a great extent, preterm birth still represents one of the most prevalent causes of death for children younger than 5 years. Infants with a birthweight of less than 1500g, referred to as very low birthweight, or infants with a birthweight less than 1000g, referred to as extremely low birthweight,6 are especially likely to face lifelong disabilities, including visual, hearing, motor, cognitive, and behavioral impairments,1 showing the need for optimisation of the clinical procedures used in the care of premature infants. Anaemia of prematurity is an issue that commonly arises in the management of premature infants,6 and is facilitated by routine blood drawing and low iron reservoirs. Since erythropoietin is mainly produced in tissues less sensitive to hypoxia in preterm infants, such as the liver, and overall erythropoietin clearance is accelerated, preterm infants show a reduced erythropoietin response in states of hypoxaemia. Moreover, the lifespan of fetal erythrocytes is lower than adult erythrocytes, which might facilitate the development of anaemia of prematurity.6 Consequently, one of two preterm infants receive red blood cell (RBC) transfusions during their hospital stay,7 the most susceptible group being infants with extremely low birthweight, with up to 90% receiving adult packed RBCs (pRBCs).6 Possible strategies to prevent the development of anaemia of prematurity constitute implementing a prudent approach to phlebotomy indications, delayed cord clamping to ensure maximal salvage of cord blood, and iron and erythropoietin supplementation to boost endogenous erythropoiesis.7 The rationale behind pRBC transfusion is to increase haematocrit and haemoglobin concentrations to offset episodes of apnea and reduce hypoxaemia7 and, consequently, to ameliorate the potential deleterious effects of anaemia of prematurity on cerebral oxygen supply.6 However, unrestrained pRBC transfusion holds some caveats. An increased incidence of adverse clinical outcomes with the transfusion of pRBCs, such as bronchopulmonary dysplasia, retinopathy of prematurity, and necrotising enterocolitis, has been a topic of scientific discourse. To our knowledge, no satisfactory conclusions have been drawn because data between most studies cannot be compared because of differences in design, quality, and heterogeneity in international and national transfusion guidelines (appendix p 2).7 We aim to summarise current evidence on the effect of pRBC transfusion on necrotising enterocolitis, broncho- pulmonary dysplasia, retinopathy of prematurity, and neurological outcomes through a comprehensive literature search (appendix p 15); elucidate patho- physiological caveats that should be considered upon transfusion of adult pRBCs into preterm infants; and discuss alternatives to previous study designs unsuitable to infer causality between pRBC transfusion and adverse clinical outcomes. Bronchopulmonary dysplasia Acute and long-term pulmonary dysfunction is a common issue in neonatology. Because of immaturity of lung tissue and impaired repair mechanisms, the pulmonary vasculature and alveoli are damaged, thus causing the development of bronchopulmonary dysplasia.2 http://crossmark.crossref.org/dialog/?doi=10.1016/S2352-3026(22)00207-1&domain=pdf e616 www.thelancet.com/haematology Vol 9 August 2022 Review Approx imately two in five infants born prematurely develop bronchopulmonary dysplasia. This number increases with decreasing gestational age, as up to 68% of infants born before a gestational age of 28 weeks are diagnosed with bronchopulmonary dysplasia. Well researched risk factors for bronchopulmonary dysplasia include prematurity, low birthweight, postnatal infections, and mechanical ventilation. It is diagnosed by a need for additional oxygen for at least 28 days postnatal and can be further classified as mild, moderate, and severe, according to the extent of the need for additional oxygen, more invasive means to deliver oxygen, and age (ie, post- menstrual age of 36 weeks [gestational ageexperience of a tertiary neonatal center in UK. J Matern Fetal Neonatal Med 2021; published online Jan 21. https://doi.org/10.1080/14767058.2021.1874910 37 Garg P, Pinotti R, Lal CV, Salas AA. Transfusion-associated necrotizing enterocolitis in preterm infants: an updated meta- analysis of observational data. J Perinat Med 2018; 46: 677–85. 38 Yeo KT, Kong JY, Sasi A, Tan K, Lai NM, Schindler T. Stopping enteral feeds for prevention of transfusion-associated necrotising enterocolitis in preterm infants. Cochrane Database Syst Rev 2019; 10: CD012888. 39 Hay S, Zupancic JAF, Flannery DD, Kirpalani H, Dukhovny D. Should we believe in transfusion-associated enterocolitis? Applying a GRADE to the literature. Semin Perinatol 2017; 41: 80–91. 40 Patel RM, Knezevic A, Shenvi N, et al. Association of red blood cell transfusion, anemia, and necrotizing enterocolitis in very low-birth- weight infants. JAMA 2016; 315: 889–97. 41 Bauer K, Linderkamp O, Versmold H. Systolic blood pressure and blood volume in preterm infants. Arch Dis Child 1993; 69: 521–22. 42 Sankaran VG, Orkin SH. The switch from fetal to adult hemoglobin. Cold Spring Harb Perspect Med 2013; 3: a011643. 43 Kaufman DP, Khattar J, Lappin SL. Physiology, fetal hemoglobin. Treasure Island, FL: StatPearls Publishing, 2021. 44 Wardle SP, Yoxall CW, Crawley E, Weindling AM. Peripheral oxygenation and anemia in preterm babies. Pediatr Res 1998; 44: 125–31. 45 Perez M, Robbins ME, Revhaug C, Saugstad OD. Oxygen radical disease in the newborn, revisited: oxidative stress and disease in the newborn period. Free Radic Biol Med 2019; 142: 61–72. 46 Gavulic AE, Dougherty D, Li S-H, et al. Fetal hemoglobin levels in premature newborns. Should we reconsider transfusion of adult donor blood? J Pediatr Surg 2021; 56: 1944–48. 47 Georgeson GD, Szőny BJ, Streitman K, et al. Antioxidant enzyme activities are decreased in preterm infants and in neonates born via caesarean section. Eur J Obstet Gynecol Reprod Biol 2002; 103: 136–39. 48 Frosali S, Di Simplicio P, Perrone S, et al. Glutathione recycling and antioxidant enzyme activities in erythrocytes of term and preterm newborns at birth. Biol Neonate 2004; 85: 188–94. 49 Hellström W, Martinsson T, Hellstrom A, Morsing E, Ley D. Fetal haemoglobin and bronchopulmonary dysplasia in neonates: an observational study. Arch Dis Child Fetal Neonatal Ed 2021; 106: 88–92. 50 Hellström W, Martinsson T, Morsing E, Gränse L, Ley D, Hellström A. Low fraction of fetal haemoglobin is associated with retinopathy of prematurity in the very preterm infant. Br J Ophthalmol 2021; 106: 970–74. 51 Teofili L, Papacci P, Orlando N, et al. Allogeneic cord blood transfusions prevent fetal haemoglobin depletion in preterm neonates. Results of the CB-TrIP study. Br J Haematol 2020; 191: 263–68. 52 Christensen RD, Jopling J, Henry E, Wiedmeier SE. The erythrocyte indices of neonates, defined using data from over 12,000 patients in a multihospital health care system. J Perinatol 2008; 28: 24–28. 53 Arbell D, Orkin B, Bar-Oz B, Barshtein G, Yedgar S. Premature red blood cells have decreased aggregation and enhanced aggregability. J Physiol Sci 2008; 58: 161–65. 54 Ruef P, Stadler AA, Poeschl J. Flow behavior of fetal, neonatal and adult RBCs in narrow (3–6 μm) capillaries—calculation and experimental application. Clin Hemorheol Microcirc 2014; 58: 317–31. 55 Kroth J, Weidlich K, Hiedl S, Nussbaum C, Christ F, Genzel-boroviczény O. Functional vessel density in the first month of life in preterm neonates. Pediatr Res 2008; 64: 567–71. 56 Genzel-Boroviczény O, Christ F, Glas V. Blood transfusion increases functional capillary density in the skin of anemic preterm infants. Pediatr Res 2004; 56: 751–55. 57 Rampling MW, Whittingstall P, Martin G, et al. A comparison of the rheologic properties of neonatal and adult blood. Pediatr Res 1989; 25: 457–60. 58 Holroyde CP, Oski FA, Gardner FH. The “pocked” erythrocyte. Red-cell surface alterations in reticuloendothelial immaturity of the neonate. N Engl J Med 1969; 281: 516–20. 59 de Porto AP, Lammers AJ, Bennink RJ, ten Berge IJ, Speelman P, Hoekstra JB. Assessment of splenic function. Eur J Clin Microbiol Infect Dis 2010; 29: 1465–73. 60 Bard H, Widness JA. The life span of erythrocytes transfused to preterm infants. Pediatr Res 1997; 42: 9–11. 61 Yoshida T, Prudent M, D’alessandro A. Red blood cell storage lesion: causes and potential clinical consequences. Blood Transfus 2019; 17: 27–52. 62 Grau M, Kuck L, Dietz T, Bloch W, Simmonds MJ. Sub-fractions of red blood cells respond differently to shear exposure following superoxide treatment. Biology (Basel) 2021; 10: 47. 63 Roussel C, Buffet PA, Amireault P. Measuring post-transfusion recovery and survival of red blood cells: strengths and weaknesses of chromium-51 labeling and alternative methods. Front Med (Lausanne) 2018; 5: 130. 64 Leal JKF, Adjobo-Hermans MJW, Bosman GJCGM. Red blood cell homeostasis: mechanisms and effects of microvesicle generation in health and disease. Front Physiol 2018; 9: 703. 65 Heaton A, Keegan T, Holme S. In vivo regeneration of red cell 2,3-diphosphoglycerate following transfusion of DPG-depleted AS-1, AS-3 and CPDA-1 red cells. Br J Haematol 1989; 71: 131–36. 66 Frank SM, Abazyan B, Ono M, et al. Decreased erythrocyte deformability after transfusion and the effects of erythrocyte storage duration. Anesth Analg 2013; 116: 975–81. 67 Xiong Y, Xiong Y, Wang Y, et al. Inhibition of glutathione synthesis via decreased glucose metabolism in stored RBCs. Cell Physiol Biochem 2018; 51: 2172–84. 68 Schaer DJ, Buehler PW, Alayash AI, Belcher JD, Vercellotti GM. Hemolysis and free hemoglobin revisited: exploring hemoglobin and hemin scavengers as a novel class of therapeutic proteins. Blood 2013; 121: 1276–84. 69 Nader E, Skinner S, Romana M, et al. Blood rheology: key parameters, impact on blood flow, role in sickle cell disease and effects of exercise. Front Physiol 2019; 10: 1329. 70 Tomaiuolo G. Biomechanical properties of red blood cells in health and disease towards microfluidics. Biomicrofluidics 2014; 8: 051501. 71 Piety NZ, Stutz J, Yilmaz N, Xia H, Yoshida T, Shevkoplyas SS. Microfluidic capillary networks are more sensitive than ektacytometry to the decline of red blood cell deformability induced by storage. Sci Rep 2021; 11: 604. 72 Antonelou MH, Seghatchian J. Insights into red blood cell storage lesion: toward a new appreciation. Transfus Apher Sci 2016; 55: 292–301. 73 Kanias T, Lanteri MC, Page GP, et al. Ethnicity, sex, and age are determinants of red blood cell storage and stress hemolysis: results of the REDS-III RBC-Omics study. Blood Adv 2017; 1: 1132–41. 74 Lanteri MC, Kanias T, Keating S, et al. Intradonor reproducibility and changes in hemolytic variables during red blood cell storage: results of recall phase of the REDS-III RBC-Omics study. Transfusion 2019; 59: 79–88. 75 Weisenhorn EM, van ‘t Erve TJ, Riley NM, Hess JR, Raife TJ, Coon JJ. Multi-omics evidence for inheritance of energy pathways in red blood cells. Mol Cell Proteomics 2016; 15: 3614–23. 76 van ‘t Erve TJ, Wagner BA, Ryckman KK, Raife TJ, Buettner GR. The concentration of glutathione in human erythrocytes is a heritable trait. Free Radic Biol Med 2013; 65: 742–49. 77 van ‘t Erve TJ, Wagner BA, Martin SM, et al. The heritability of metabolite concentrations in stored human red blood cells. Transfusion 2014; 54: 2055–63. Copyright © 2022 Elsevier Ltd. All rights reservedlung injury, leading to the corrosion of pulmonary capillary endothelium, and resulting in rapidly progressing pulmonary oedema.12 Whether transfusion-associated lung injury seen in adults can also be observed in preterm infants has not been clarified. Nevertheless, the concept of neonatal post- transfusion lung injury has been investigated and discussed by Rashid and colleagues.12 Furthermore, both prospective8 and retrospective studies9–11 were able to find associations between pRBC transfusion and the development of bronchopulmonary dysplasia. Timing and volume of transfusions were also associated with the development of bronchopulmonary dysplasia.10 For example, in one study (appendix p 4), the group receiving transfusion before 3 weeks of age had a significantly higher incidence of bronchopulmonary dysplasia than the group receiving late transfusion (ie, at 3 weeks of age or later). Another study (appendix p 4) found an association between the use of multiple pRBC transfusions and increased incidence of broncho- pulmonary dysplasia; however, the authors did not disclose the criteria according to which they initiated transfusions (appendix p 8). Concerning the number of transfusions, in a retrospective study on infants with extremely low birthweight, the authors showed an association between pRBC transfusions, and number of pRBC transfusions, and the diagnosis of bronchopulmonary dysplasia11 at 28 days of age. Furthermore, Patel and colleagues found a connection between total volume of administered pRBCs and incidence of bronchopulmonary dysplasia in infants with very low birthweight. The study was also able to show an association between oral intake of iron supplementation and incidence of bronchopulmonary dysplasia, and the authors discuss a possible connection to reactive oxygen species formation; however, there are not enough data available to sufficiently support this association. As for the pRBC transfusion, again, transfusion criteria used in the study population were not specified.10 The scarcity of disclosure concerning transfusion practices in studies evaluating the effect of pRBC transfusion on the risk of bronchopulmonary dysplasia is both common and hinders the comparison of published data. Most studies do not properly correct for known risk factors of broncho- pulmonary dysplasia, such as oxygen supplementation, nor do they sufficiently report their method in the manuscript, which increases the risk for random effects.10,11 Compounding this issue, the definition of broncho pulmonary dysplasia also slightly deviates between studies, as exemplified by Valieva and colleagues,11 since the authors defined bronchopulmonary dysplasia as the need for oxygen at a postnatal age of 28 days and at a corrected gestational age of 36 weeks. By contrast, another study defined broncho pulmonary dysplasia solely by the need for oxygen supplementation at a postmenstrual age of 36 weeks.9 Regarding transfusion thresholds used in studies that disclosed details on transfusion criteria, further discrepancies can be noted (table 1, appendix p 4). Valieva and colleagues11 used haematocrit thresholds of 20%, 30%, and 35%, depending on clinical features, such as how much oxygen the patient needed, or whether the patient was mechanically ventilated or showed cardiorespiratory stability. By contrast, other studies allowed transfusion to be initiated at the discretion of the neonatologist,9 or did not disclose any transfusion criteria at all in their methods.10 There are also differences concerning the age and composition of blood bags, as one study allowed the blood aliquots to be as old as 42 days when transfused.11 Another study did not use pRBCs that had been divided into aliquots at all.9 Moreover, only a few studies provided information on composition and preparation of pRBCs,8,10,11 and inclusion criteria varied considerably across different studies. For example, whereas most studies had 1500 g as the upper weight limit,8–10 one study exclusively included infants with extremely low birthweight with a weight ranging from 500 g to 1000 g.11 Furthermore, other studies defined the developmental stage suitable for inclusion with the weight at birth rather than gestational age.8–10 Altogether, these factors make it difficult to merge the available data into a comprehensive summary on the effect of pRBC transfusion on the incidence of bronchopulmonary dysplasia. Adverse neurological outcome Prematurity commonly leads to neurological deficits. Up to 30% of all preterm infants who survive show www.thelancet.com/haematology Vol 9 August 2022 e617 Review H ae m at oc rit H ae m og lo bi n Ve nt ila ti on Cl in ic al fe at ur es Tr ea tm en t o f pR BC s St or ag e m ed iu m Tr an sf us io n vo lu m e Du ra ti on o f p RB C st or ag e Do no r ch ar ac te ris ti cs Bl au e t a l ( 20 11 )13 Co ho rt st ud y, n= 36 2 8 da ys a nd re tic ul oc yt e co un t 0· 4 O r s ym pt om s o f A O P, h yp ov ol em ia , m aj or su rg er y Le uk or ed uc ed a nd irr ad ia te d CP D an d SA GM 10 –2 0 m L/ kg in 4 h W ith in 5 d ay s o f co lle ct io n N A As a bo ve N A 10 g /d L N A Cr iti ca lly il l ( un de fin ed ) Le uk or ed uc ed a nd irr ad ia te d CP D an d SA GM 10 –2 0 m L/ kg in 4 h W ith in 5 d ay s o f co lle ct io n N A Sh ah e t a l15 Ra nd om ise d co nt ro lle d tr ia l; n= 99 ≤1 8% ≤6 g /d L N A As ym pt om at ic an d re tic ul oc yt e co un t 60 p er m in ), an d do ub lin gof o xy ge n re qu ire m en t i n th e pr ev io us 4 8 h 24 h ta ch yc ar di a (> 18 0 pe r m in ), la ct at e ≥2 ·5 m q/ L o r m et ab ol ic ac id os is (p H 1 20 k ca l/k g/ d, su rg er y w ith in 2 4 h N A N A 20 m L/ kg N A N A As a bo ve ≤2 5% ≤8 g /d L An y m ec ha ni ca l v en til at io n w ith F iO 2 ≤ 0· 4; o r C PA P >6 cm H 2O a nd F IO 2 ≥ 0· 4 N A N A N A N A N A N A As a bo ve ≤3 0% ≤1 0 g/ dL M od er at e ve nt ila tio n (M AP >8 cm H 2O a nd F iO 2 > 0· 4) N A N A N A 15 m L/ kg N A N A N o lib er al v er su s r es tr ict iv e t ra ns fu sio n th re sh ol d st ud ie s w er e in clu de d in th is ta bl e. H ae m at oc rit a nd h ae m og lo bi n va lu es a re co ns id er ed a s c ut off v al ue s; th e lo w es t v al ue sh ou ld b e t he lo w er m ar gi n, b ut if a dd iti on al fa ct or s a re p re se nt (e g, n ec es sit y of v en til at io n) , t he n ex t h ig he st cu to ff m ar ks th e lo w er m ar gi n fo r w he n to in iti at e t ra ns fu sio n. N A= no t a va ila bl e. A DS O L= ad en in e d ex tr os e so rb ito l s od iu m ch lo rid e m an ni to l. A DS O L- 3= AD SO L ve rs io n N ut ric el . A DS O L- 5= AD SO L ve rs io n O pt iso l. AO P= an ae m ia o f p re m at ur ity . C PA P= co nt in uo us p os iti ve a irw ay p re ss ur e. C PD =c itr at e ph os ph at e d ex tr os e. F iO 2= fra ct io na l c on ce nt ra tio n of o xy ge n in in sp ire d ai r. M AP =m ea n ar te ria l p re ss ur e. N IV =n on -in va siv e ve nt ila tio n. p RB Cs =p ac ke d re d bl oo d ce lls . S AG M =s od iu m a de ni ne g lu co se m an ni to l. +C rit er ia th at m us t b e fu lfi lle d ad di tio na lly . Ta bl e 1 : T ra ns fu si on -r el at ed ch ar ac te ris ti cs o f p ro sp ec ti ve st ud ie s o n pR BC tr an sf us io ns in p re te rm in fa nt s r ep or ti ng e ff ec ts o n co m pl ic at io ns o f p re m at ur it y e618 www.thelancet.com/haematology Vol 9 August 2022 Review neurosensory or neuromotor impairments.16 Neuro- development is most frequently assessed with the third edition of the Bayley Scales of Infant and Toddler Development (BSID),15,17 and cerebral palsy and gross motor functions are quantified by the Gross Motor Function Classification System.16 Clinical trials that assessed the effect of pRBC transfusion on various neurodevelopmental outcomes have provided conflicting results. Most of the trials compared different transfusion thresholds, either favouring a liberal18 or restrictive19 approach to transfusion, depending on higher or lower haematocrit threshold levels triggering transfusion, or identifying no differences between the two strategies.3,16,17,20 Only a few studies evaluated the overall effect of pRBC transfusion on neurodevelopmental outcomes.15,21,22 Although a randomised controlled trial by Bell and colleagues18 found a higher incidence of periventricular leukomalacia or intraparenchymal brain haemorrhage in the restrictive transfusion group of infants with very low birthweight and extremely low birthweight, Whyte and colleagues16 showed no beneficial effect of maintaining high or low haemoglobin thresholds. Only after adjusting the Mental Development Index score cutoff for cognitive delay from less than 70 to less than 85, they were able to show a significant improvement in the liberal transfusion group, thus favouring higher transfusion thresholds. However, a study that followed up 44 children for more than a decade, who were initially included in the study collective of Bell and colleagues,18 compared intracranial volume with healthy controls. The study found that children given pRBC transfusion showed a reduction in brain volume, with a tendency for the highest reduction in children transfused under a liberal regimen. Nonethe- less, it should be mentioned that the number of total transfusions did not differ between the groups (p=0·15),19 challenging the interpretability of the results.19 However, a third study with the same collective showed reductions in neurocognitive outcomes, such as associative verbal fluency, visual memory, and reading in infants transfused under liberal guidelines 13 years after the start of the first study.23 Moreover, two studies identified neurocognitive development and outcome at the age 18–26 month follow- up to be inversely correlated with total transfusion count, transfusion donor exposure, and transfusion volume.15,24 The heterogeneity in results is further exacerbated by a retrospective analysis showing a positive correlation between the number of transfusions in the first 7 days of life and a more favourable neurocognitive outcome at the age 18–24 month follow-up in infants with extremely low birthweight;21 a Cochrane review with low risk of bias showing no difference in neurocognitive outcome between liberal and restrictive threshold transfusion in infants with very low birthweight after 18–22 months of follow-up;25 and a systematic review and meta-analysis showing no advantage of any transfusion threshold for neurodevelopmental outcome in infants with very low birthweight (table 2).26 Transfusion thresholds varied between trials (table 1, appendix p 4). For example, whereas some trials defined thresholds based solely on the method of O2 supplementation,18 others additionally used distinct numeric values for fractional concentration of oxygen in inspired air (FiO2)15,21 and positive expiratory pressure15 to guide their interventions. In intubated patients, Bell and colleagues18 used 46% haematocrit in blood as the liberal threshold and 34% as the restrictive threshold, whereas Shah and colleagues transfused patients under respiratory support at 31% haematocrit.15 Transfusion volume was also not standardised, varying between 10 mL/kg and 20 mL/kg.15,18,21 Moreover, some studies chose transfusion volume according to the patient’s respiratory status and clinical presentation.15 Regarding blood handling, studies did not consistently report storage conditions, storage solution, or whether irradiation was done. For example, whereas Bell and colleagues18 included no information about pretreatments and storage, Wang and colleagues21 simply mentioned to have used leukocyte-poor concentrates, and Shah and colleagues15 reported the use of irradiated and leukocyte- depleted erythrocyte concentrates treated with citrate- phosphate-dextrose adenine. In addition to the studies on infants with very low birthweight, several randomised controlled trials investigating the effect of transfusion thresholds on neurological or neurocognitive outcomes in infants with extremely low birthweight have been done, of which none showed a significant association.3,16,17,20 Although these trials are of high methodological quality, their comparability is limited by the absence of an international consensus on optimal transfusion guidelines. Although the cutoff for transfusion in the liberal group of the ETTNO trial20 for infants up to 7 days of postnatal age was set at 41% haematocrit for critical individuals and 35% haematocrit for non-critical individuals, the TOP trial17 by Kirpalani and colleagues included haematocrit thresholds of 38% and 35%, depending on the use of respiratory support. Differences are also observable in the restrictive group, with transfusion thresholds set at 34% and 28% haematocrit in the ETTNO trial20 versus 32% and 29% haematocrit in the TOP trial.17 Moreover, follow-up length varied between18 months and 26 months.16,17,20 Studies also differed in the volume of pRBCs transfused, with the ETTNO trial20 using a transfusion volume of 20 mL/kg, and Kirpalani and colleagues using 15 mL/kg bodyweight.17 The chosen neurodevelopmental scores also differed between the studies; some studies used the second edition of the BSID,16,21 other studies used the third edition of the BSID,15,17 and one study combined the second and third edition of the BSID.20 Of note, the quality of methods and statistical power of the ETTNO trial (n=1013)20 and TOP trial (n=1824)17 provide robust evidence for equivalency of low and high transfusion threshold strategies in infants with extremely www.thelancet.com/haematology Vol 9 August 2022 e619 Review St ud y ty pe Cl in ic al fe at ur es M ea su re m en t/ In te rv en ti on N eu ro de ve lo pm en ta l ev al ua ti on N eu ro de ve lo pm en ta l o ut co m e Au th or s’ co nc lu si on Vu e t a l ( 20 21 );24 n= 62 8 Po st -h oc a na ly sis of a ra nd om ise d co nt ro lle d tr ia l Ge st at io na l a ge 24 –2 7 w ee ks As so cia tio n be tw ee n pR BC tr an sf us io n an d ne ur od ev el op m en ta l ou tc om e, fo llo w -u p fo r 22 –2 6 m on th s BS ID -II I s co re Δ- sc or e w ith e ac h tr an sf us io n: –0 ·9 6 (9 5% C I – 1· 34 to –0 ·5 7) , pl v ol um e m ea su re m en ts in di ca te d ad ju st ed m ea ns Ki rp al an i e t a l ( 20 06 );3 (n =4 51 ) Ra nd om ise d co nt ro lle d tr ia l EL BW a nd ge st at io na l a ge actually differed in transfusion protocols, and only one study explicitly reported the number of transfusions received within the first 10 days of life.3,17,20 Considering the pathophysiology of retinopathy of prematurity and the data collated by Lust and colleagues,27 the number of immediate postnatal transfusions might influence retinopathy of prematurity development and be distorted by the time needed for patient allocation. Storage conditions and the handling of pRBCs were poorly documented in terms of pretreatments (irradiation, leukoreduction, storage solution).11,27,30,32 Moreover, hetero- geneity also arises from patient selection. Although some groups characterised retinopathy of prematurity in infants with extremely low birthweight,26 other studies focused on infants with very low birthweight,30 or disregarded common classifications altogether,29 which is particularly problematic because both birthweight and gestational age are well known risk factors for the development of retinopathy of prematurity.4 Because of low comparability between retrospective studies, reduced focus on retinopathy of prematurity in large prospective trials that disregard early transfusion and focus on infants with extremely low birthweight, and low sample size of prospective trials that evaluated the relationship between retinopathy of prematurity and transfusion, the association between pRBC transfusion and retinopathy of prematurity remains unclear. Since the molecular pathophysiology of retinopathy of prematurity and the effects of pRBC transfusion are highly intertwined, however, cause–effect correlation seems plausible. Thus, future prospective studies focusing on retinopathy of prematurity with standardised methods are warranted. Necrotising enterocolitis Another complication associated with prematurity, especially concerning infants with very low birthweight, is necrotising enterocolitis. Up to 7% of preterm infants in the USA and Canada are diagnosed with necrotising enterocolitis, up to 30% of whom do not survive.5 Symptoms are food intolerance, abdominal distension, and haemato chezia. The pathophysiology of necrotising enterocolitis revolves around immaturity of the gastrointestinal tract and concomitant inadequacy of intestinal barrier function. Gastrointestinal microbial dysbiosis and dysregulated immune response also contribute to the risk of necrotising enterocolitis.5 It has been postulated that necrotising enter ocolitis is triggered as a consequence of pRBC transfusions. Referred to as transfusion-associated necrotising entero colitis (TANEC), possible patho physio logical mech anisms include hypoxia- induced intestinal damage exacerbated by pRBC transfusion, also termed hypo perfusion-reperfusion injury.6 There is also accumu lating evidence for a possible connection between trans fusion and the development of necrotising entero colitis,13,33–35 and worse clinical outcomes in TANEC.36 Blau and colleagues13 found an association between the timing of pRBC transfusion and necrotising enterocolitis in some infants with very low birthweight. They reported that patients with TANEC tended to have lower haema- tocrit values and weighed less than the non-transfusion- associated necrotising enterocolitis group. By contrast, Mally and colleagues35 found that preterm births in the TANEC group were better developed and healthier than in the non-transfusion-associated group, suggesting that pRBC transfusion might have induced development of necrotising enterocolitis. The two studies also differed in reported duration to the onset of TANEC (5 h [SD 1·0] vs 22 h [SD 5·0]).13,35 Observational studies did not find any association between transfusion and the development of necrotising enterocolitis.21,37 Thus, the association remains unclear, as supported by a meta-analysis of observational studies that remained inconclusive on the relation between transfusion and necrotising enterocolitis.37 Moreover, there are considerable differences concerning inclusion, exclusion, and transfusion criteria between individual studies. For example, some studies chose an upper weight limit of 1500 g instead of an upper age limit.32 Regarding the definition of necrotising enterocolitis, most chose Bell’s stage II as threshold,34,36 whereas other studies used a slightly different definition (Bell’s stage IIb).13 Another point in which study protocols differed was enteral feeding. A randomised trial comparing doppler ultrasound flow examinations of the superior mesenteric artery between fed and fasting infants during transfusion showed that transfusions reduce postprandial increase in mesenteric blood flow velocity, especially in infants weighing more than 1250 g.14 The results correspond with the findings of Mally and colleagues,35 who concluded that TANEC is associated with the intake of full enteral meals. Nevertheless, a meta- analysis remained inconclusive because of the scarcity of sufficient high-quality data.38 Furthermore, in some studies, enteral feeding was stopped during transfusion,33 but was not stopped,34 or not disclosed in other studies.21 Indications to initiate pRBC transfusion differed notably between studies (appendix p 4). Blau and colleagues13 initiated transfusion upon five or six episodes of apnea within 8 h, or if preterm infants did not gain e622 www.thelancet.com/haematology Vol 9 August 2022 Review 10–15 g/kg bodyweight per day within a week, or had lethargy, unexplained tachycardia or tachypnea, insufficient reticulocyte count increase, or a haematocrit less than 25%. Patients recruited in a study by Wang and colleagues,21 however, were transfused at a haematocrit less than 21% and no reticulocyte response if asymptomatic; at a haematocrit less than 31% if symptoms of inadequate oxygenation, such as the need for oxygen or ventilation and not gaining weight, were present; and at a haematocrit less than 36% upon need for more than 35% oxygen or ventilator assisted breathing, or upon phlebotomy blood loss of 15% of total blood volume within the first 7 days of life. One study prescribed concomitant furosemide infusion,14 whereas in other studies no such procedure was mentioned.13 Moreover, both rate and duration of transfusions differed, with one study transfusing 10–15 mL/kg within 2–3 h,21 and a set of studies transfusing 15 mL/kg within 4 h.13,34 By contrast, another study provided detailed transfusion criteria and a formula to calculate transfusion volume. The timeframe in which pRBCs were administered, however, remained unspecified.33 Of note, only a few studies provided information on the blood bag itself,13,21,24 such as its age or expiration date.13 In accordance with a systematic review by Hay and colleagues,39 the differences in study protocols impede data comparison. More transparency and conscientiousness concerning handling and storage of pRBC packs and transfusion practices are warranted. This need holds especially true considering hypoperfusion–reperfusion injury as a possible cause of necrotising enterocolitis. Of note, it is severe anaemia, not transfusions, that is associated with the development of necrotising enterocolitis.40 Genetic background and microbial composition5 substantially influence the risk for necrotising enterocolitis. Carefully designed protocols in future should also consider the potential association between donor characteristics and the occurrence of necrotising enterocolitis. Physiological aspects: adult cells in a premature body Approximately 90% of infants with extremely low birthweight and 40% of infants with very low birthweight6 receive at least one pRBC transfusion. Since top-up transfusions for non-bleeding individuals are usually done at 15 mL/kg,7 and blood volume of preterm infants weighing 800–1499 g approximates 83 mL/kg,41 15–20% of total blood volume is added with each pRBC transfusion. Consideringboth, heterogeneity in biophysical and biochemical properties between different blood packages (interdonor variability and inherent pRBC quality) and differences in neonatal vasculature and erythrocyte physiology between adults and neonates, the complexity of blood transfusion for preterm infants exceeds the scope of conventional clinical studies, calling for additional translational and basic research. Fetal haemoglobin differs from that in adults. The first developmental switch, the replacement of embryonic haemoglobin with fetal haemoglobin occurs within the first trimester when erythropoiesis relocates from the yolk sack to the liver.42 The second switch, a process unique to higher primates,42 takes place after birth and is completed within the first 6 months of life. Fetal haemoglobin shows higher affinity to oxygen and lower affinity to 2,3-dipho- sphoglycerate than adult haemoglobin, resulting in a left shift of the oxygen dissociation curve and a reduced ability to release oxygen in the periphery.43 Therefore, transfusion of adult pRBCs containing adult haemoglobin might lead to a sudden increase in the availability of oxygen, thus facilitating oxygen delivery to peripheral tissues.44 This process might be of pathophysiological relevance because of increased formation of ROS, potentially resulting in subsequent tissue damage.45 This hypothesis is complemented by the findings that (1) fetal haemoglobin concentrations of extremely low gestational age preterm infants transiently drop by 30% after one pRBC transfusion and by 60% after two pRBC transfusions, possibly because of dilution;46 (2) the activity of antioxidant enzymes, such as catalase, glutathione peroxidase, or superoxide dismutase, is reduced in preterm infants, as measured in cord blood hemolysates;47 (3) erythrocytes of neonates have higher glutathione turnover-rates than adult erythrocytes;48 and (4) low fetal haemoglobin concen- trations were associated with the development of broncho- pulmonary dysplasia49 and retino pathy of prematurity.50 To counteract these findings, more physiological approaches, such as cord-blood-derived pRBC transfusion, have been proposed. One study, although limited by small sample size, showed cord blood pRBC transfusion to circumvent sudden drops in fetal haemoglobin.51 RBCs of premature infants not only differ in haemoglobin composition but also in morphology and flow behaviour. Erythrocytes of preterm infants born with a gestational age of between 22 weeks and 32 weeks present with a mean corpuscular volume between 110 fL and 120 fL and a mean corpuscular haemoglobin concentration between 37 pg and 40·5 pg compared with adult values of 80–94 fL for mean corpuscular volume and 27–31 pg for mean corpuscular haemoglobin concentration.52 This increased volume could account for the enhanced probability of aggregation and rouleaux formation, as measured in plasma free medium in vitro.53 Moreover, the increased cell volume of about 120 fL and the concomitant need for additional deformability54 might necessitate the larger lumen of capillaries, arterioles, and venules in preterm neonates, with diameters of 7–24 µm being reported.55,56 However, rouleaux formation and overall blood viscosity is reduced in neonates, especially under low shear stress.57 This finding is accounted for by differences in plasma composition and reduced protein content, which counteract the enhanced probability of aggregation and rouleaux formation and decreases the risk of capillary obstruction.54 In summary, the www.thelancet.com/haematology Vol 9 August 2022 e623 Review physiological challenges induced by the transfusion of adult pRBCs into preterm infants and how they affect the development of unfavourable outcomes remain unclear, showing the need for additional studies (figure). Blood of premature infants contains a high abundance of so-called pitted (or pocked) erythrocytes (30–87%) when compared to adults (3%).58 Their number is inversely correlated to gestational age and birthweight and has been shown to be a reliable marker of splenic hypofunction or immaturity. These findings are suggestive of impaired splenic cell clearance in premature infants,59 which is of particular interest, since the life span of pRBCs transfused to preterm infants is approximately 60 days,60 compared with 120 days in adults.7 Therefore, extrasplenic clearance or erythrolysis might account for the accelerated cell loss after transfusion.60 The differences in erythrocyte and circulatory physiology between adults and preterm infants warrant further selection criteria for pRBCs used in neonatal transfusion. Biochemical and biophysical erythrocyte functionality declines over time during cold storage, resulting in collectively termed storage lesions.61 Large differences in erythrocyte stress resistance and functionality already become apparent immediately after blood donation and processing,62 but are not considered in blood banking. Storage lesions have a negative effect on the number of cells remaining in circulation 24 h after transfusion. A loss of 25% is currently deemed acceptable,63 and the mean average loss of cells 24 h after transfusion is 15%.63 Erythrocytes undergoing lysis or retained in the spleen not only delay reaching the desired haematocrit as calculated from the volume of transfused pRBCs, but are also a burden to the patient because of the release of deleterious cell components, such as microvesicles and free or oxidised haemoglobin.64 During routine cold storage, metabolites in pRBCs undergo changes in abundance, as extensively documented in a review incorporating data obtained from omics technologies, in addition to available literature.61 All approved storage solutions contain adenine as a building block for ATP and glucose to maintain chemical energy production. The decline of adenine and glucose, both intracellularly and in the storage medium, is rapidly reverted once pRBCs re-enter the blood, thus deemed unproblematic at first glance.65 However, other storage lesions, such as reduced shear stress and oxidative stress resistance and deformability, are irreversible.66 Most of these storage lesions are caused by (oxidative) damage to membrane components, since defence against reactive oxygen species becomes compromised during storage.67 Because of pleiotropic deleterious effects, free haemoglobin, haem, and iron might be of concern,68 particularly in preterm infants. Moreover, ageing pRBCs increasingly shed microvesicles that continue to be suspected as being problematic in vivo, especially if their concentration in the blood is suddenly increased after transfusion.64 At least three storage lesions have been connected to post-transfusion erythrocyte recovery so far: cell morphology, intracellular ATP content, and deformability.63 Evaluation of all three parameters is within the limits of standard technology, thus their measurement could be implemented at various degrees of effort. Analysis of morphology, especially for echinocytes and spherocytes, requires simple cell type counting in a microscope; ATP concentration can be measured with commercially available test kits; and the elongation index, the most important rheological property related to cell deformability,69 could be measured with a microfluidic slit-flow ektacytometer.70 Ektacytometry would be a larger investment and require more qualified personnel to operate than other options, however, studies have shown that new methods for evaluating deformability of red blood cells with microfluidic capillary networks are more sensitive than ektacytometry for detecting storage induced changes in red blood cell theological properties, thus enabling faster data acquisition and routine monitoring of pRBC quality in clinical routine.71 Contrary to transfusion practice in adults, in which more than one pRBC unit is applied most of thetime, more extensive characterisation of erythrocytes seems feasible and warranted in preterm infants, in whom less than one unit is needed. Besides the option to evaluate outcome related cell properties ahead of transfusion, erythrocytes from different donors vary in a plethora of biological parameters that can make them more suitable or less suitable for transfusion.72 The variability between donors, although adequately evaluated in biochemical studies, has not been evaluated in clinical trials and is, therefore, not routinely recorded with standard clinical tests in blood banks. For example, storage haemolysis, and osmotic and oxidative Figure: Adult erythrocytes in an immature body Adult red blood cells differ substantially from the red blood cells of premature infants. 0 25 50 75 100 O 2 s at ur at io n of h ae m og lo bi n (% ) Partial pressure of O2 Mean corpuscular volume 80–94 fL 110–120 fL Adult haemoglobin Fetal haemoglobin Adults Premature infants Adults Premature infants 4–6 μm 8–14 μm Haemoglobin Reference viscosity Plasma proteins: 60–80 g/L Fibrinogen: 300–400 mg/dL Reduced viscosity Plasma proteins: 30-50 g/L Fibrinogen: 120-250 mg/dL Life span of adult packed red blood cells: 120 days Normal clearance: 3% pitted red blood cells Life span of adult packed red blood cells: 56 days Extrasplenic clearance: 30–87% pitted red blood cells Capillary diameter Plasma composition Clearance 15–20% of blood volume HbF HbA ? e624 www.thelancet.com/haematology Vol 9 August 2022 Review stress resistance of pRBCs have been associated with donor age, sex, and ethnicity.73 Moreover, osmotic and mechanical stress resistance, as well as oxidative haemolysis to some extent, have been reported to be donor specific.74 Compounding the findings on donor variability, identical twin studies have shown that glycolysis (both enzyme activity and metabolite abundance), pentose–phosphate pathway activity, and in consequence, glutathione concentrations, are heritable.75,76 Importantly, these results were also reproduced in stored pRBCs according to the work by van ’t Erve and colleagues.77 Thus, erythrocyte function seems to underly interindividual and heritable differences even on a molecular scale. Interindividual differences in pRBC quality and clinical studies Although a plethora of retrospective analyses, cohort studies, and case-control studies have reported pRBC transfusion to be associated with adverse outcomes in preterm infants, high-quality RCTs focusing on differences between liberal versus restrictive transfusion thresholds have not reproduced such results. Paradox at first glance, the controversy between the results of RCTs and cohort studies, supported by absence of standardisation and limited comparability even between similarly designed studies, might be an overall consequence of not accounting for interindividual pRBC quality in clinical trials. As retrospective, cohort, and case-control studies compare transfused individuals with non-transfused individuals, they have an inherent bias: transfused individuals are a-priori sicker than non- transfused. Although studies account for this bias by applying multivariate analyses to correct correlations for known risk factors, they tend to overestimate effect sizes (ie, adverse association between pRBC transfusion and outcome). The problem of a-priori sickness has been circumvented with the introduction of liberal versus restrictive transfusion studies, as preterm infants are randomly assigned into either one of the groups. Usually, infants randomly assigned to the liberal group are more likely to be transfused and receive an overall higher volume of pRBCs. Yet, considering sample sizes, differences in transfusion incidence between groups might not be large enough to account for the overestimated effects observed in cohort studies, leaving observable adverse clinical outcomes dependent on differences in overall transfusion number and volume. Therefore, the analysis of a group of mostly transfused individuals holds caveats because donor exposure usually is not conclusively reported or does not differ between groups; donor data are neither recorded nor reported; and interindividual differences in pRBC quality are not quantified ahead of transfusion. Molecular erythrocyte properties are heritable, making pRBCs more suitable or less suitable for transfusion. The tolerability of pRBC transfusion might be shaped by differences in donor RBC quality rather than volume or number of transfusions. Thus, adverse clinical outcomes might be insufficiently recorded in liberal versus restrictive transfusion studies because overall erythrocyte quality equilibrates between large treatment groups that consist mainly of transfused individuals. The hypotheses that we have discussed would offer an explanation as to why studies evaluating the effect of transfusion versus no transfusion find associations between administration of pRBCs and adverse clinical outcomes in preterm infants, whereas restrictive versus liberal transfusion threshold studies do not find any association. Conclusions Differences in neonatal transfusion guidelines hinder the generation of comparable clinical data, and heterogeneity in clinical data prevents the implementation of standardised transfusion protocols in preterm infants. Although cohort studies have reported associations between pRBC transfusion and adverse outcomes in preterm infants, RCTs investigating different transfusion thresholds did not find such correlations. To provide Search strategy and selection criteria This is a comprehensive Review of clinical studies on packed red blood cell (pRBC) transfusion, neurocognitive outcomes, and the risk for bronchopulmonary dysplasia, necrotising enterocolitis, and retinopathy of prematurity. For comparisons between studies, we included and comprehensively discussed original articles, systematic reviews, and meta-analyses in the text. We only further screened and summarised full articles in English with relevance to the topic in descriptive tables. We also retrieved numeric values, such as means, standard deviations, and sample sizes, and presented them in tables. A post-hoc cross-validation systematic research was done by the information retrieval office of the university library of the Medical University of Vienna, and 578 articles published in Ovid MEDLINE until April 27, 2022, were screened. From these articles, 224 were marginally related to the topic of pRBC transfusion and its association with necrotising enterocolitis, bronchopulmonary dysplasia, retinopathy of prematurity, and neurodevelopmental outcomes. Of these articles, we considered 19 studies relevant. We divided the research strategy into sections and their combinations. The main sections considered relevant words referring to: population of interest, procedure, outcomes, and type of study. We chose articles to be included into full-text analysis according to their overall methodological (sample acquisition and selection, randomisation, and if applicable, sample size) and statistical quality (correction for risk factors, multivariate models), or according to their representativeness for methodological flaws (if cited). We organised the literature according to prospective and retrospective designs and favoured prospective trials for inclusion. www.thelancet.com/haematology Vol 9 August 2022 e625 Review causal rather than correlational evidence for the association between pRBC transfusion and adverse outcomes in the clinically vulnerable group of preterm infants, novel study designs that consider interindividual differences in pRBC quality are warranted. Therefore, longitudinal prospective studies with a clinical– biochemical transla tional design might finally circum- vent this causality problem. Contributors ME had the idea for the manuscript. ME, LB, US,EWM, VG, and AB conceptualised the manuscript. LB, ME, VG, EWM, and AB prepared the original draft. LB, ME, VG, EWM, US, DS, AH, and AB wrote the manuscript and LB, ME, VG, EWM, US, LW, AB, AR, and KK-S reviewed and edited it. VG, LW, and AB were project supervisors for the neonatology section. US and EWM were project supervisors for the physiology section. All authors have read and agreed to the submitted version of the manuscript. Declaration of interests We declare no competing interests. Acknowledgments We thank the mentoring programme of the Medical University of Vienna (Vienna, Austria), for facilitating cooperation between different departments involved in this manuscript. We also thank Brigitte Wildner, from the information retrieval office at the university library of the Medical University of Vienna, for supporting the research strategy of this Review with extreme professionality and competence. References 1 March of Dimes, Partnership for Maternal, Newborn and Child Health, Save the Children, WHO. Executive summary. In: Howson C, Kinney M, Lawn J, eds. Born too soon: the global action report on preterm birth. Geneva: World Health Organization; 2012: 1–7. 2 Kalikkot Thekkeveedu R, Guaman MC, Shivanna B. Bronchopulmonary dysplasia: a review of pathogenesis and pathophysiology. Respir Med 2017; 132: 170–77. 3 Kirpalani H, Whyte RK, Andersen C, et al. The Premature Infants in Need of Transfusion (PINT) study: a randomized, controlled trial of a restrictive (low) versus liberal (high) transfusion threshold for extremely low birth weight infants. J Pediatr 2006; 149: 301–07. 4 Cavallaro G, Filippi L, Bagnoli P, et al. The pathophysiology of retinopathy of prematurity: an update of previous and recent knowledge. Acta Ophthalmol 2014; 92: 2–20. 5 Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011; 364: 255–64. 6 Cibulskis CC, Maheshwari A, Rao R, Mathur AM. Anemia of prematurity: how low is too low? J Perinatol 2021; 41: 1244–57. 7 Villeneuve A, Arsenault V, Lacroix J, Tucci M. Neonatal red blood cell transfusion. Vox Sang 2021; 116: 366–78. 8 Ghirardello S, Dusi E, Cortinovis I, et al. Effects of red blood cell transfusions on the risk of developing complications or death: an observational study of a cohort of very low birth weight infants. Am J Perinatol 2017; 34: 88–95. 9 Lee EY, Kim SS, Park GY, Lee SH. Effect of red blood cell transfusion on short-term outcomes in very low birth weight infants. Clin Exp Pediatr 2020; 63: 56–62. 10 Patel RM, Knezevic A, Yang J, et al. Enteral iron supplementation, red blood cell transfusion, and risk of bronchopulmonary dysplasia in very-low-birth-weight infants. Transfusion 2019; 59: 1675–82. 11 Valieva OA, Strandjord TP, Mayock DE, Juul SE. Effects of transfusions in extremely low birth weight infants: a retrospective study. J Pediatr 2009; 155: 331–37.e1. 12 Rashid N, Al-Sufayan F, Seshia MMK, Baier RJ. Post transfusion lung injury in the neonatal population. J Perinatol 2013; 33: 292–96. 13 Blau J, Calo JM, Dozor D, Sutton M, Alpan G, La Gamma EF. Transfusion-related acute gut injury: necrotizing enterocolitis in very low birth weight neonates after packed red blood cell transfusion. J Pediatr 2011; 158: 403–09. 14 Krimmel GA, Baker R, Yanowitz TD. Blood transfusion alters the superior mesenteric artery blood flow velocity response to feeding in premature infants. Am J Perinatol 2009; 26: 99–105. 15 Shah P, Cannon DC, Lowe JR, et al. Effect of blood transfusions on cognitive development in very low birth weight infants. J Perinatol 2021; 41: 1412–18. 16 Whyte RK, Kirpalani H, Asztalos EV, et al. Neurodevelopmental outcome of extremely low birth weight infants randomly assigned to restrictive or liberal hemoglobin thresholds for blood transfusion. Pediatrics 2009; 123: 207–13. 17 Kirpalani H, Bell EF, Hintz SR, et al. Higher or lower hemoglobin transfusion thresholds for preterm infants. N Engl J Med 2020; 383: 2639–51. 18 Bell EF, Strauss RG, Widness JA, et al. Randomized trial of liberal versus restrictive guidelines for red blood cell transfusion in preterm infants. Pediatrics 2005; 115: 1685–91. 19 Nopoulos PC, Conrad AL, Bell EF, et al. Long-term outcome of brain structure in premature infants: effects of liberal vs restricted red blood cell transfusions. Arch Pediatr Adolesc Med 2011; 165: 443–50. 20 Franz AR, Engel C, Bassler D, et al. Effects of liberal vs restrictive transfusion thresholds on survival and neurocognitive outcomes in extremely low-birth-weight infants: the ETTNO randomized clinical trial. JAMA 2020; 324: 560–70. 21 Wang Y-C, Chan O-W, Chiang M-C, et al. Red blood cell transfusion and clinical outcomes in extremely low birth weight preterm infants. Pediatr Neonatol 2017; 58: 216–22. 22 Fontana C, Raffaeli G, Pesenti N, et al. Red blood cell transfusions in preterm newborns and neurodevelopmental outcomes at 2 and 5 years of age. Blood Tranfus 2022; 20: 40–49. 23 McCoy TE, Conrad AL, Richman LC, Lindgren SD, Nopoulos PC, Bell EF. Neurocognitive profiles of preterm infants randomly assigned to lower or higher hematocrit thresholds for transfusion. Child Neuropsychol 2011; 17: 347–67. 24 Vu PT, Ohls RK, Mayock DE, et al. Transfusions and neurodevelopmental outcomes in extremely low gestation neonates enrolled in the PENUT Trial: a randomized clinical trial. Pediatr Res 2021; 90: 109–16. 25 Whyte R, Kirpalani H. Low versus high haemoglobin concentration threshold for blood transfusion for preventing morbidity and mortality in very low birth weight infants. Cochrane Database Syst Rev 2011; 11: CD000512. 26 Wang P, Wang X, Deng H, et al. Restrictive versus liberal transfusion thresholds in very low birth weight infants: a systematic review with meta-analysis. PLoS One 2021; 16: e0256810. 27 Lust C, Vesoulis Z, Jackups R Jr, Liao S, Rao R, Mathur AM. Early red cell transfusion is associated with development of severe retinopathy of prematurity. J Perinatol 2019; 39: 393–400. 28 Zhu Z, Hua X, Yu Y, Zhu P, Hong K, Ke Y. Effect of red blood cell transfusion on the development of retinopathy of prematurity: a systematic review and meta-analysis. PLoS One 2020; 15: e0234266. 29 Brooks SE, Marcus DM, Gillis D, Pirie E, Johnson MH, Bhatia J. The effect of blood transfusion protocol on retinopathy of prematurity: a prospective, randomized study. Pediatrics 1999; 104: 514–18. 30 Chen H-L, Tseng H-I, Lu C-C, Yang S-N, Fan H-C, Yang R-C. Effect of blood transfusions on the outcome of very low body weight preterm infants under two different transfusion criteria. Pediatr Neonatol 2009; 50: 110–16. 31 Knee D, Knoop S, Davis AT, Rawson B, DiCarlo A, Olivero R. Outcomes after implementing restrictive blood transfusion criteria in extremely premature infants. J Perinatol 2019; 39: 1089–97. 32 Del Vecchio A, Henry E, D‘Amato G, et al. Instituting a program to reduce the erythrocyte transfusion rate was accompanied by reductions in the incidence of bronchopulmonary dysplasia, retinopathy of prematurity and necrotizing enterocolitis. J Matern Fetal Neonatal Med 2013; 26 (suppl 2): 77–79. 33 Teišerskas J, Bartašienė R, Tamelienė R. Associations between red blood cell transfusions and necrotizing enterocolitis in very low birth weight infants: ten-year data of a tertiary neonatal unit. Medicina (Kaunas) 2019; 55: 16. 34 Wan-Huen P, Bateman D, Shapiro DM, Parravicini E. Packed red blood cell transfusion is an independent risk factor for necrotizing enterocolitis in premature infants. J Perinatol 2013; 33: 786–90. e626 www.thelancet.com/haematology Vol 9 August 2022 Review 35 Mally P, Golombek SG, Mishra R, et al. Association of necrotizing enterocolitis with elective packed red blood cell transfusions in stable, growing, premature neonates. Am J Perinatol 2006; 23: 451–58. 36 Stokes V, Rajai A, Mukherjee D, Mukherjee A. Transfusion- associated necrotizing enterocolitis (NEC) in extremely preterm infants: