Background Hyperkalemic cardiac arrest is a potential complication of massive transfusion in children. Our objective was to identify risk factors and potential preventive measures by reviewing the literature on transfusion‐associated hyperkalemic cardiac arrest ( TAHCA ) in the pediatric population. Study Design and Methods Literature searches were performed in MEDLINE and the C ochrane D atabase of S ystematic R eviews. Results We identified nine case reports of pediatric patients who had experienced cardiac arrest during massive transfusion. Serum potassium concentration was reported in eight of those reports; the mean was 9.2 ± 1.8 mmol/ L . Risk factors for TAHCA noted in the case reports included infancy (n = 6); age of red blood cells (RBCs; n = 5); site of transfusion (n = 5); and the presence of comorbidities such as hyperkalemia, hypocalcemia, acidemia, and hypotension (n = 9). We also identified 13 clinical studies that examined potassium levels associated with transfusion. Of those 13, five studied routine transfusion, two were registries, and six examined massive transfusion. Conclusions Key points identified from this literature search are as follows: 1) Case reports are skewed toward infants and neonates in particular and 2) the rate of blood transfusion, more so than total volume, cardiac output, and the site of infusion, are key factors in the development of TAHCA . Measures to reduce the risk of TAHCA in young children include anticipating and replacing blood loss before significant hemodynamic compromise occurs, using larger‐bore (>23‐gauge) peripheral intravenous catheters rather than central venous access, checking and correcting electrolyte abnormalities frequently, and using fresher RBCs for massive transfusion.
Extremely high viremic levels of parvovirus B19 (B19V) can be found in acutely infected, but asymptomatic donors. However, reports of transmission by single-donor blood components are rare. In this prospective study, paired donor-recipient samples were used to investigate the transfusion risk.Posttransfusion plasma or blood samples from recipients were tested for B19V DNA by polymerase chain reaction, generally at 4 and 8 weeks, and for anti-B19V immunoglobulin (Ig)G by enzyme immunoassay, at 12 and 24 weeks. To rule out infection unrelated to transfusion, pretransfusion samples and linked donor's samples for each B19V DNA-positive recipient were assayed for B19V DNA and anti-B19V IgG and IgM. To confirm transmission, sequencing and phylogenetic analysis were performed.A total of 14 of 869 (1.6%) recipients were B19V DNA positive, but only 1 of 869 (0.12%; 95% confidence interval, 0.0029%-0.6409%) was negative for B19V DNA and anti-B19V IgG before transfusion and seroconverted posttransfusion. This newly infected patient received 5 × 10(10) IU B19V DNA in one red blood cell (RBC) unit from an acutely infected anti-B19V-negative donor in addition to RBCs from three other donors that cumulatively contained 1320 IU of anti-B19V IgG. DNA sequencing and phylogenetic analysis showed that sequences from the linked donor and recipient were identical (Genotype 1), thus establishing transfusion transmission.The 0.12% transmission rate documented here, although low, could nonetheless result in hundreds or thousands of infections annually in the United States based on calculated confidence limits. Although most would be asymptomatic, some could have severe clinical outcomes, especially in neonates and those with immunocompromised or hemolytic states.
Among older children with sickle cell anemia, leukocyte counts, hemoglobin, and reticulocytosis have previously been suggested as disease severity markers. Here we explored whether these blood parameters may be useful to predict early childhood disease severity when tested in early infancy, defined as postnatal ages 60-180 days.Data from fifty-nine subjects who were followed at Children's National Medical Center's Sickle Cell Program for at least three years was retrospectively analyzed. Comparisons were made between white blood cell counts, hemoglobin and reticulocyte levels measured at ages 60-180 days and the clinical course of sickle cell anemia during infancy and childhood.A majority of subjects had demonstrable anemia with increased reticulocytosis. Only increased absolute reticulocyte levels during early infancy were associated with a significant increase in hospitalization during the first three years of life. Higher absolute reticulocyte counts were also associated with a markedly shorter time to first hospitalizations and a four-fold higher cumulative frequency of clinical manifestations over the first three years of life. No significant increase in white blood cell counts was identified among the infant subjects.These data suggest that during early infancy, increased reticulocytosis among asymptomatic SCA subjects is associated with increased severity of disease in childhood.
Leukoreduction to eliminate mononuclear cells within blood products is necessary to prevent graft-versus-host disease after transfusion. Published reports document low concentrations of mononuclear cells leftover in fresh-frozen plasma products, however the phenotype and the proliferative potential of these cells has not been tested.We investigated residual cellular components contained within fresh and fresh-frozen plasma products and characterised their proliferative potential in co-cultures with unrelated allogeneic cells. We designed a flow-based assay to phenotype cells and quantify cell division by measuring the dilution of fluorescently labeled protein as cells divide. Leukocytes from consenting donors were purified from fresh liquid or fresh-frozen plasma units and cultured for three to seven days with unrelated irradiated allogeneic targets.We discovered a median of 1.6×107 viable lymphocytes were detectable in fresh plasma units after collection (n=8), comprised of a mixture of CD3+ CD8+ and CD3+ CD4+ cells. Furthermore, we identified a median of 8.4% of live CD3+ plasma lymphocytes divided as early as Day 4 when co-cultured with unrelated allogeneic cells, expanding to a median 88.8% by Day 7 (n=3). Although freezing the plasma product reduced the total number of viable leukocyte cells down to 2.3×105 (n=10), residual naive CD3+ cells were viable and demonstrated division through Day 7 of co-culture.The evidence of viable proliferative lymphocytes in fresh and fresh-frozen plasma products derived from centrifugation suggests that additional leukoreduction measures should be investigated to fully eradicate reactive lymphocytes from centrifuged plasma products.
AccessScience is an authoritative and dynamic online resource that contains incisively written, high-quality educational material covering all major scientific disciplines. An acclaimed gateway to scientific knowledge, AccessScience is continually expanding the ways it can demonstrate and explain core, trustworthy scientific information that inspires and guides users to deeper knowledge.
Hemolytic disease of the newborn occurs as a result of sensitization of the mother's immune system to red-cell antigens of the fetus. This sensitization results from the transplacental passage of fetal red cells possessing an antigen that is not present on maternal red cells. The IgG antibodies produced by the mother in response to the foreign antigen cross the placenta into the fetal circulation, bind to the red cells there, and destroy them. As a consequence of the hemolytic process, there is both extramedullary hematopoiesis and reticuloendothelial clearance of sensitized fetal cells, leading to hepatic and splenic enlargement. As the . . .
T activation is the earliest known and the most common form of polyagglutination, a group of conditions characterized by alterations in red blood cell (RBC) membrane glycoprotein structure in certain pathological states, resulting in the agglutination of such transformed cells in the presence of most ABO compatible adult sera. In vitro polyagglutination was first described by Hübener (1925) and then by Thomsen (1927), but it was Thomsen's graduate student, Friedenreich (1930), who elucidated the underlying mechanism and named it T haemagglutination, after his mentor. The alteration of red-cell membrane structure is simply the removal of N-acetyl neuraminic acid residues from the normally disialylated tetrasaccharides of the MN, Ss and other RBC membrane sialoglycoproteins by the enzymatic action of bacterial neuraminidase. Cleavage of the neuraminic acid residues exposes the hidden beta-linked galactosyl residue (Gal-β(1–3)-GalNAc), the T cryptantigen (Fig 1). The unmasked T antigen binds with the anti-T IgM antibodies that are normally present in human adult plasma, resulting in agglutination in vitro and possible haemolysis in vivo. The phenomenon is usually transient, often lasting only a few days or weeks, rarely persisting for months (Obeid et al, 1977). Schematic representation of tetrasaccharide and paragloboside of red blood cell membrane glycoproteins wIth modified structures expressing T, Tn and Tk receptors. Other variants of T activation secondary to bacterial enzyme action on red-cell antigens are less common. Th transformation is considered to be an incomplete or intermediate form of T activation, as further action by stronger neuraminidases results in the classic T activation. Tk polyagglutination is as a result of microbial β-galactosidases that cleave a galactose residue from paragloboside, exposing N-acetylglucosamine, the Tk receptor. Transient Tx activation has been described in children with pneumococcal infection and also in a child with acute haemolytic anaemia, but no identifiable infection, but the molecular mechanism of this condition has not been clarified (Wolach et al, 1987). Tn activation is different from other forms of T activation in that it is persistent and is not associated with microbial enzyme action, but is the result of mutation in a clone of red cells, with the abnormal cells containing a cryptantigen with only one acetylated galactose residue without the beta-linkage, secondary to the absence of the enzyme 3-β-galactosyltransferase (Cartron et al, 1978). Tn polyagglutination has been associated with myelodysplastic syndromes and acute leukaemia in adults (Ness et al, 1979). In newborn infants, however, transient Tn polyagglutination has been described, often associated with an antenatal flu–like illness in the mother, and is probably as a result of delayed maturation or temporary deficiency of β-3-galactosyltransferase in the fetus (Rose et al, 1983). Variants of T activation and polyagglutination because of congenital or acquired alterations in non-T RBC surface antigens (Cad, VA, acquired B, haemoglobin M-Hyde Park, etc.) are not discussed further in this review (Horn, 1999). In clinical practice, T-antigen exposure is most often associated with neuraminidase-producing organisms, particularly with Clostridium perfringens and Streptococcus pneumoniae. Other organisms implicated include Bacteroides, Escherichia coli and other gram-negative bacilli, such as Actinomyces and influenza virus. Vibrio Cholerae produces neuraminidase and T activation in vitro;in vivo T activation secondary to this organism has not been reported in the English literature. The degree of T activation of red cells may be influenced by the amount of neuraminidase in the circulation and is probably mitigated in part by serum neuraminidase inhibitors. T-transformed red cells may be destroyed quickly by immune-mediated intravascular haemolysis, secondary to complement-fixing anti-T IgM antibodies. In addition, the T-activated red cells may be cleared more rapidly from the circulation because of a paucity of membrane sialic acid residues (Durocher et al, 1975). Anti-T IgM antibodies are absent at birth and in early infancy, but are present in most adults. The formation of these antibodies has been attributed to antigenic stimulation by intestinal flora that possess antigens similar to T antigens, analogous to the development of ABO antibodies secondary to environmental antigenic stimuli (Springer & Tegtmeyer, 1981). Contrary to earlier haemagglutination methods that failed to detect any anti-T activity in umbilical cord sera, sensitive radioimmunoassay techniques have been able to demonstrate the presence of small quantities of anti-T antibodies of the non-agglutinating IgG class in cord sera, probably of maternal origin (Kim, 1980). The ubiquitous nature of anti-T antibodies in adults would suggest that exposure of the T antigen would result in autoagglutination and haemolysis, but this has been rarely demonstrated (Moores et al, 1975), and indeed, most reports on T activation do not indicate whether anti-T antibody is present in the patient's serum during the acute disease. The absence of anti-T in the plasma of adults with T-activated red cells has been variously attributed to adsorption of antibody to the red cells or to immune tolerance or immune paralysis (Issit, 1985). There are isolated case reports of acute intravascular haemolysis in adult patients with sepsis and T activation, resulting in precipitous drops in the haematocrit to levels as low as 5–12%, but, in at least some of the cases, the persistence of anti-T antibody indicated that T activation was not the culprit, but bacterial lecithinases were responsible (Judd et al, 1982: Hubl et al, 1993). Latent T-antigen receptors are present on leucocytes and platelets too, but induction of T activation of platelets in vitro does not seem to alter platelet function significantly (Hysell et al, 1976). T activation is rare in the normal population, with only one case being identified in a random screening of 10 000 hospital patients (Rawlinson & Stratton, 1984). In hospitalized adults considered to be at high risk for red-cell polyagglutination because of malignancy and/or sepsis, 18 out of 238 patients (7·6%) were found to have cryptantigen exposure, with 13 of the 18 patients being classified as T-activated on the basis of lectin panel tests (Buskila et al, 1987) A similar prevalence rate of 7% has been noted in patients with acquired immunodeficiency syndrome, a group of patients who are subject to severe and recurrent infections with organisms known to cause red-cell membrane modifications (Adams et al, 1989). In adult surgical intensive care patients with septicaemia, RBC T activation has been demonstrated in 32%, with 72% of such patients having evidence of infection with bacteria with the ability to release neuraminidase. Furthermore, 71% of patients with T activation had elevated free-serum haemoglobin levels compared with 14% of patients without T activation (Lenz et al, 1987). Invasive Streptococcus pneumoniae infection is being increasingly recognized as an important cause of atypical haemolytic uraemic syndrome (HUS) in children (Cabrera et al, 1998; McTaggart & Burke, 1998; Shirey et al, 1999). Circulating neuraminidase produced by the pneumococci has been postulated to play a pathogenetic role in this syndrome, causing exposure of the T cryptantigen in renal endothelium, red cells and platelets with subsequent binding of anti-T IgM, resulting in the characteristic triad of renal failure, microangiopathic haemolytic anaemia and thrombocytopenia (Klein et al, 1977). Exposure of the T cryptantigen in the glomeruli has been demonstrated by fluoroscein-labelled peanut lectin in renal biopsy specimens. The transfusion of plasma products containing anti-T antibody could be expected to lead to acceleration of haemolysis of T-activated red cells, but it may be difficult to differentiate this from the ongoing haemolysis that is characteristic of HUS and has not been clearly demonstrated in many reports. Earlier studies reported mortality rates as high as 50%; decreased mortality rates in recent years have been attributed to early recognition of this condition and the judicious use of blood products (McGraw et al, 1989; Erickson et al, 1994). Neonates with necrotizing enterocolitis (NEC) have been considered to be especially susceptible to T activation and its resulting complications. NEC is the most common gastrointestinal emergency in the intensive care nursery, predominantly affecting premature neonates. This clinical syndrome, characterized by abdominal distention, feeding intolerance, occult or gross blood in the stools, together with systemic signs ranging from lethargy and temperature instability, to shock, metabolic acidosis and disseminated intravascular coagulation, afflicts an estimated 1200–9600 infants in the United States annually, with a fatality rate of 9–28% (Stoll, 1994). Although the pathogenesis of the disease is not clear, bacterial infection of the intestine appears to be an important risk factor, at least in a subset of infants with the disease. Fulminant NEC has been associated with Clostridial infection, but recent prospective studies have shown no difference in bacterial colonization of the gut in neonates with NEC compared with age and gestation-matched controls (Peter et al, 1999). Exposure of the neonatal red cells to locally released bacterial neuraminidase from devitalized bowel has been postulated to result in the T activation that is reported in 11–27% of infants with NEC (Klein et al, 1986; Williams et al, 1989; Novak, 1990; Kirsten et al, 1996; Osborn et al, 1999). Klein et al (1986) detected T activation in 17 (27%) out of 62 infants with NEC, but only in two (0·001%) out of 1600 infants with other illnesses admitted to a neonatal intensive care unit. Infants with NEC who were T-activated were four times more likely to require surgical intervention than infants who were not. Clostridia were cultured from the blood, peritoneal fluid or stool in 88% of the infants with T activation. Other authors have also noted the association between the incidence and strength of T activation and severity of NEC, with strongly T-activated infants often manifesting symptoms of fulminant NEC, with intestinal perforation or gangrene requiring operative intervention, and a higher mortality (Kirsten et al, 1996; Osborn et al, 1999). The association with anaerobic infection or neuraminidase-producing bacteria is not always clearly demonstrated or reported. Enteric anaerobes have sometimes been identified in T-activated infants with NEC only in post-mortem peritoneal cultures, although repeated cultures prior to death were negative (Williams et al, 1989; Marshall et al, 1993). T activation has also been reported in a large percentage of neonates with bowel-related surgical problems (other than NEC) and sepsis (Grant et al, 1998). Earlier reports indicated that the transfusion of adult blood containing anti-T to infants with T activation resulted in severe and sometimes fatal intravascular haemolysis, and indeed, routine screening of infants with NEC for T activation has been recommended by some authors to avoid transfusion-related morbidity and mortality (Rodwell & Tudehope, 1993; Kirsten et al, 1996). Again, it is not clear from examination of the available evidence that T activation alone was responsible for the haemolysis in many of these infants. Williams et al (1989) described haemolysis in four out of six T-activated infants who received plasma containing blood products, mild or no haemolysis in five patients who received low anti-T-titre plasma and no haemolysis in T-activated infants who received no plasma containing blood products. Klein et al (1986) noted haemolysis in three out of four T-activated infants who received plasma containing blood products, but haemolysis occurred in only one out of 13 infants who received only washed blood products. A recent report from a neonatal intensive care unit in Australia on a large series of infants with NEC, showed that evidence of moderate to severe haemolysis was present in seven out of 15 T-activated infants, despite the use of fresh-frozen plasma with a low titre of anti-T (Osborn et al, 1999). The necessity for routine screening and special transfusion protocols for infants with NEC is highly controversial. A recent international forum showed considerable variation in practice in centres around the world (Engelfriet et al, 1999). Two out of eight transfusion centres routinely screened infants with NEC for polyagglutination, two centres screened selected cases only if there was evidence of haemolysis and no screening was done at all in the other centres as no cases of haemolysis as a result of T activation had been identified for years. Obviously, the centres that did not believe that T activation in infants was a major problem also did not feel that special transfusion protocols were necessary for these infants. Blood-bank and transfusion-service procedures have made it more difficult to recognize T activation in the laboratory in recent years, particularly in neonates, so the problem of T activation may well be underestimated unless clinicians are vigilant. T activation should be suspected in patients at risk who have evidence of intravascular haemolysis with haemoglobinuria and haemoglobinaemia following transfusion of blood products or unexplained failure to achieve the expected post-transfusion haemoglobin increment. Polyagglutinability of T-transformed red cells is not seen when monoclonal anti-A and anti-B reagents are used for ABO grouping. Conventional cross-matching and compatibility testing beyond the determination of blood group is considered to be unnecessary and is not routinely performed before transfusing infants less than 4 months of age, as the formation of alloantibodies is extremely rare (Voak et al, 1994; Menitove, 1999). T activation may be suspected in the laboratory if there are inconsistencies in blood group and type over time, if there are disagreements in forward and reverse groupings or if a minor cross-match of the patient's red cells with donor serum reveals agglutination (Kumar & Sethi, 1993). Unlike normal RBCs that aggregate when mixed with positively charged substances, such as hexadimethrene bromide (Polybrene), T-activated red cells often fail to aggregate, because desialysation results in a net reduction of negative charge at the RBC surface (Issitt, 1985). The sensitivity of this test is dependent on the degree of desialysation and the proportion of T-activated cells in the sample. The diagnosis may be confirmed by testing with a lectin panel, where the affinity of plant lectins for specific sugars on the cell surface permit differentiation of the variants of polyagglutination (Table I). It must be pointed out that these tests, although sensitive, simple and inexpensive, may be difficult to standardize. Arachis hypogea anti-T lectin is simply a saline extract of processed raw peanuts (Vengelan-Tyler, 1999); activity levels of the reagent vary with the seeds used, the extraction methods and the age of the reagents. Commercial lectin panels are available. Transfusion protocols that have been advocated for T-activated patients include the use of washed or plasma-reduced red blood cells and platelets, and the use of low anti-T-titre plasma if fresh-frozen plasma is essential. RBCs preserved in extended storage media from which nearly all plasma has been removed may be considered for infants needing multiple small-volume transfusions in order to reduce donor exposures (Engelfriet et al, 1999). There is no clear definition of 'low titre' anti-T for screening blood donors. Low titre has been variously defined as titres not greater than 1 in saline (Engelfriet et al, 1999), agglutination titres less than 8 (Issitt et al, 1972) or an optical density value of 0·300 at 405 nm using a haemolysis test (Lynen et al, 1991). A 'slow' transfusion with close evaluation of possible haemolysis has also been recommended, with measurements of haemoglobin, haptoglobin, total and indirect bilirubin, and urine haemoglobin before, during and after the transfusion (Chambers & Green, 1996). Careful selection of donor plasma units by using a minor cross-match allows transfusion of T-activated patients without untoward reactions (Eversole et al, 1986). In infants with severe haemolysis resulting in severe hyperbilirubinaemia, exchange transfusion using washed packed-red cells resuspended in additive–preservative solutions or albumin has been recommended to decrease bilirubin levels and prevent further haemolysis by replacing a significant percentage of T-activated red blood cells with donor cells not expressing the cryptantigen, but this procedure may be associated with an unacceptably high mortality in critically ill premature infants (Williams et al, 1989; Squire et al, 1992; Osborn et al, 1999). The large number of reports that describe the temporal association of haemolysis with blood transfusion in T-activated patients suggest that this phenomenon is more than a laboratory oddity. However, T activation alone may not be an adequate explanation of haemolysis in some critically ill individuals, in whom haemolysis may be secondary to bacterial lecithinases or disseminated intravascular coagulation. The condition may be a marker and not the cause of severity of illness in patients with anaerobic or pneumococcal sepsis, but assessment of the prognostic implications is undermined by the lack of adequate controls in most studies. In addition, the variation in recognition of this phenomenon and the relative importance given to it in different institutions is also remarkable. The quandary of management of the sick T-activated patient will probably remain uncertain, given the logistical and ethical problems in designing an adequate study to examine the necessity of using low-titre anti-T plasma in this situation. The diagnostic, prognostic and therapeutic implications of T activation demand awareness among clinicians, better methods to aid accurate diagnosis and exclude other causes of haemolysis, and measures to allow selection of appropriate transfusion support.
