Search

BACKGROUND: TRALI is a serious adverse effect of blood transfusion. There is evidence that the condition is underrecognized and underreported. STUDY DESIGN AND METHODS: This study was an observational study carried out in a single hospital. RESULTS: Eleven cases of TRALI were recognized over 12 years. In 10 cases the implicated donor unit was FFP and in 1 case uncertain. All implicated donors were parous women. In 4 cases the presumed causative antibodies were to an HLA class II antigen only. Specific anti-neutrophil antibodies, possibly causative, were detected in 1 case only. Ten of the 11 cases required mechanical ventilatory support. Five persons died as a result of the TRALI. The observed incidence of TRALI caused by FFP is 1 in 7900 units transfused. CONCLUSION: TRALI is the most common serious adverse effect of blood transfusion in our hospital. Antibodies to HLA class II antigens should be looked for routinely when investigating a possible case of TRALI.

BACKGROUND: Cytotoxic WBC antibodies are found in patients who have refractoriness to platelet transfusion (RPT) or are experiencing febrile transfusion reactions (FTRs) and in sera giving so called nonspecific hemagglutination by IAT (N/S IAT). Sera from such patients were screened for WBC antibodies regardless of the ability to fix complement using a flow cytometric (FC) lymphocyte indirect immunofluorescence test (LIFT) to compare FC-LIFT with a routine lymphocytotoxicity test (LCT) for WBC antibody detection. STUDY DESIGN AND METHODS: Serum from 104 patients with RPT, 87 with FTR, and 147 with N/S IAT were tested in parallel by using FC-LIFT and LCT. Sera giving discrepant results were re-tested with an HLA class I antibody ELISA to assess whether they were HLA-specific. RESULTS: Of the sera tested, 175 were LIFT positive, and 146 were LCT positive. Fifty-five had antibodies that were detectable only by LIFT; 26 were positive only by LCT. Of these 81 discrepant sera, 30 of 63 were positive in HLA ELISA. CONCLUSION: FC-LIFT detects more WBC antibodies than does LCT or ELISA, and it is a superior screening technique. Because some cytotoxic antibodies are detectable only by LCT, comprehensive WBC antibody screening would require the application of both techniques. However, because FC assessments of cytotoxicity have been described, LCTs may become redundant for WBC antibody screening.

G. L. Daniels (Chair), D. J. Anstee, J. P. Cartron, W. Dahr, A. Fletcher, G. Garratty, S. Henry, J. Jorgensen, W. J. Judd, L. K ornstad, C. Levene, M. Lin, C. Lomas-Francis, A. Lubenko, J. J. Moulds, J. M. Moulds, M. Moulds, M. Overbeeke, M. E. Reid, P. Rouger, M. Scott, P. Sistonen, E. Smart, Y. Tani, S. Wendel & T. Zelinski*

BACKGROUND: GPB carries ‘N’ at its N-terminus and S and s, determined by a polymorphism at amino acid position 29 (Met29Thr). The low-incidence antigens Mv, sD, and Mit are associated with weakened expression of S and/or s, and the purpose of this study was to define their molecular bases. METHODS: The GPB gene (GYPB) was sequenced after RT-PCR of RNA from four samples: two Mv+, one sD+, and one Mit+. The point mutations observed were confirmed by sequencing of genomic DNA from these and other examples of sD+ and Mit+ samples. RESULTS: A point mutation of 65C>G observed in the Mv+ samples predicted a change of Thr3Ser. A mutation of C>G at nucleotide 173 of the GYPB coding sequence, observed in two sD+ samples, predicted a change of Pro39Arg. Three Mit+ samples showed a nucleotide substitution of 161G>A, which predicted a change of Arg35His. Altered expression of S or s was confirmed by serologic tests. CONCLUSION: These results confirm that Arg35 is important for full expression of S. Pro39 and, surprisingly, Thr3 are also important for full expression of s. Furthermore, Thr3 must be essential for expression of ‘N,’ as Mv+ RBCs lack ‘N.’

Autoimmune phenomena are well-recognised complications of Waldenstrom's macroglobulinemia (WM) and IgM monoclonal gammopathy. Peripheral neuropathy and cold agglutinin hemolytic anemia are the most common reported and occur in 5-10% of patients. Autoimmune thrombocytopenia has been rarely reported in WM and its incidence is not known. In this study we report the case of a 67-year-old man who presented with autoimmune thrombocytopenia who was subsequently found to have WM. Laboratory investigation demonstrated that platelet-associated IgM (PAIgM) but not PAIgG was clearly elevated compared to normal controls. In addition the patient's serum reacted strongly with a panel of donor platelets analysed with an indirect platelet immunofluorescence assay utilising an anti-IgM secondary antibody. Glycoprotein specificity could not however be demonstrated by ELISA techniques for platelet glycoproteins IIbIIIa, IaIIa, IbIXa, and IV. We also reviewed the case records of 104 additional cases of WM diagnosed at our institution between 5/93 and 5/99. Three further cases with clinically significant autoimmune thrombocytopenia were identified. The overall incidence of autoimmune thrombocytopenia (4/105, 3.8%) in this cohort of patients was similar to the incidence of peripheral neuropathy (7/105, 6.7%) and cold agglutinins (3/105, 2.9%).

Peripheral blood progenitor cells used during high dose treatments for malignancy may be contaminated with tumour cells that could later contribute to recurrence. CD34+ selected harvests still contain tumour cells and an additional negative selection may be capable of reducing this contamination. We have assessed a two-stage technique in which a CD34+ selection is followed by a tumour specific depletion stage using a B cell or breast cancer specific antibody panel. Initial small-scale selections on 11 patients with NHL and breast cancer showed that cell loss was greatest following the CD34+ selection with a median yield of 38.8 per cent (range 17. 2-56.4 per cent). The addition of the depletion stage resulted in a minimal loss of CD34+ cells with a yield for this step of 94.2 per cent (range 77.5-99.3 per cent). Clinical scale selections were performed on seven patients with CLL and a median of 2.8x10(6)/kg CD34+ cells (range 1.5-6.1x10(6)/kg) were collected. Cell recovery was 53.3 per cent following CD34+ selection and 76.9 per cent following the tumour specific depletion stage, resulting in a final product containing a median of 1.0x10(6)/kg CD34+ cells (range 0. 55-2.0x10(6)/kg). All unmanipulated harvests were heavily contaminated with tumour cells (median contamination 10.2 per cent, range 2.0-83.1 per cent) as measured by flow cytometry and a median 4.7 log (range 3-5 log) tumour cell purge was produced following two-stage selection. Six of the patients have received cells manipulated in this way with median engraftment times of neutrophils>0.5x10(9)/l=16 days (range 13-20 days) and platelets>20x10(9)/l=16.5 days (range 11-42 days). At a median follow-up of 25 months, these transplanted patients remain well and in molecular complete remission.

Anti-RhD immunoglobulin was administered to RhD-negative women based on estimates of fetal bleed size obtained using a direct immunofluorescence flow cytometric technique employing a FITC-conjugated monoclonal human anti-D (BRAD 3). The effectiveness of the dose administered was assessed by (i) measuring the fraction of RhD-positive fetal cells in the maternal circulation at d0, and between d2 and d10 post RhD Ig administration, (ii) quantifying the amount of anti-D detectable in maternal plasma following RhD Ig injection in the perinatal period and (iii) assessing maternal serum for the presence of immune anti-D in follow-up samples taken 3 months to 3 years after delivery. Fifty-four women were assessed, 29 having fetal bleeds in excess of 4 mL. Follow-up samples were received from 20/29 mothers after RhD Ig administration; 43-99% and 69-99% of fetal cells had been cleared by d2/3 and d5/6, respectively, in 14/20 mothers, whereas less than 50% had been cleared in the remaining mothers. Long-term follow-up samples were obtained from eight of the 29 mothers (four with bleeds >/=20 mL, two with bleeds >95 mL): none had detectable anti-D in the serum 4 months to 3 years after delivery despite the persistence of up to 36% fetal RhD-positive cells in the maternal circulation six days after delivery.

BACKGROUND: Noncomplement-fixing white cell antibodies have been demonstrated by the use of immunofluorescence flow cytometry against intact lymphocytes. However, such antibodies may be either HLA-specific or directed against other white cell antigens. Commercial enzyme-linked immunosorbent assay (ELISA) kits, using solubilized HLA molecules as targets, enable such HLA-specific antibodies to be detected in patients who are refractory to platelet transfusion, patients experiencing febrile transfusion reactions, and patients whose sera give nonspecific hemagglutination in indirect antiglobulin tests. STUDY DESIGN AND METHODS: Sera from all three groups of patients, previously screened for cytotoxic antibodies by using complement-dependent lymphocytotoxicity, were re-investigated with commercial ELISA kits for HLA antibody screening and identification using the manufacturers' recommended test methods. RESULTS: Non-complement fixing HLA antibodies were detected by ELISA in many sera that were lymphocytotoxicity test- negative; that is, 14 (17.5%) of 80 from refractory patients, 8 (23.5%) of 34 from those with febrile reactions, and 11 (22.4%) of 49 from those with nonspecific hemagglutination in the direct antiglobulin test. However, not all cytotoxic white cell antibodies were detectable by ELISA: only 19 (82.6%) of 23, 19 (67.8%) of 28, and 11 (73.6%) of 49, respectively in the three groups. Similarly, only 143 (79.4%) of 181 cytotoxic sera with clear-cut HLA-A or -B locus specificities were detectable by ELISA. CONCLUSION: ELISAs detect some but not all clinically significant HLA antibodies, irrespective of their ability to fix complement in vitro.

Background: The ready availability of red cells of the Miltenberger (Mi) class III phenotype (6.28%) prompted the study of Mi antibodies among Chinese blood donors in Hong Kong, 98 percent of whom are descended from inhabitants of Guangdong Province in southern China. Study Design and Methods: Red cells of the Mi class III phenotype were used to conduct a survey of the frequency of Miltenberger antibodies in 56,161 random Chinese blood donors, over a period of 12 months, using a microplate technique. Results: Sera from 32 donors (0.057%) were found to contain Mi antibodies: sera from 22 contained anti-Mur + Hut; sera from 4 contained anti-Vw + Mur + Hut; sera from 4 had monospecific anti- Mur; and sera from 2 had monospecific anti-Hil. The immunoglobulin isotypes of 24 sera were mixtures of IgM and IgG, 4 were pure IgM, and 4 were pure IgG. Conclusion: The majority of Mi antibodies detected were naturally occurring. This survey proved useful for mass screening of random donors for the procurement of valuable Mi antisera.

To demonstrate that fetal human platelet antigen (HPA1) type can be determined, without the need for fetal blood sampling, by amplification of fetal DNA from amniotic fluid cells using polymerase chain reaction and allele specific oligonucleotide hybridisation.Oligonucleotide DNA primers were designed to amplify a portion of the platelet glycoprotein GpIIIa gene which spans the site of the single base change which differentiates HPA1a from HPA1b. Specific oligonucleotides were designed to hybridise either to the amplified HPA1a allele or to the HPA1b allele. Amniotic cells were used as the DNA template both directly and following formal isolation of DNA. Fetal HPA1 type, determined by this method in fifteen pregnancies not at risk of perinatal alloimmune thrombocytopaenia, was compared to typing of fetal blood obtained following cordocentesis. The methodology was then used to HPA type the fetus in two pregnancies at risk of the disease.Department of Molecular Biology and Centre for Fetal Care, Queen Charlotte's Hospital.Fifteen women undergoing amniocentesis and fetal blood sampling for other indications and two women at risk of perinatal allo-immune thrombocytopaenia whose partners were heterozygotes.In the 15 control cases and the two clinical cases, determination of fetal HPA1 type from amniotic fluid cells agreed with typing of fetal blood. There was no difference in the efficiency of amplification from amniotic fluid cells directly or from isolated DNA.Fetal HPA type may be reliably determined by amplification of DNA from amniotic fluid cells, eliminating the need for fetal blood sampling or immunoglobulin administration when the fetus is HPA1a negative.

D. J . Anstee, G. W. G. Bird E. Brodheim, J.-P. Cartron W. Dahr, C. P. Engelfriet P. D. Issitt, J . Jgrgensen L. Kornstad, M . Lewis C. Levene, A . Lubenko D. Mallory, P. Morel R . Nordhagen, Y. Okubo M . Reid, Ph. Rouger Ch. Salmon, S. Seidl, I? Sistonen S. Wendel, G. Woodfield 7: Zelinski International Society of Blood Transfusion Socihte lnternationale de Transfusion Sanguine Section Editor: B. Habibi, Paris

Genetic studies of the families of two unrelated propositi prove that the low-incidence antigen ELO (ISBT No. 700.51) is not part of the MNS, Rh, Duffy, Kidd, Xg, Kx and Gerbich blood group systems and indicate that it is probably not part of the Lutheran, Scianna, Dombrock or Colton systems. Serological studies indicate exclusion from the Kell and Chido/Rodgers systems. Although the ELO antigen is rare, anti-ELO is not uncommon in sera containing at least one specificity for a low-incidence antigen.

We report an antibody (anti-TSEN) that recognizes an antigen (TSEN) at the unique amino acid sequence that results from the junction of GPA58 to GPB27 if the GPB carries the S antigen. Red cells from several unrelated donors that possess this specific GP(A-B) hybrid molecule were agglutinated by anti-TSEN. Since a synthetic peptide with the amino acid sequence at this junction (Pro-Glu-Glu-Glu-Thr-Gly-Glu-Met-Gly-Gln-Leu-Val-His-Arg) specifically inhibited anti-TSEN, it must detect an antigen within this novel amino acid sequence. The TSEN antigen has been provisionally assigned the MNS blood group system number 002.033 on behalf of the ISBT Working Party on Terminology for Red Cell Surface Antigens.

A 35-year-old Brazilian woman (gravida 4, para 2) was delivered of a severely anaemic child whose cord red blood cells had a strongly positive direct antiglobulin test and who required two exchange transfusions within 24 h of birth. Because of the emergency of the situation and the lack of a local immunohaematology reference laboratory, the phenotype of the mother and the specificity of the relevant antibody could not be determined. Hence, compatible blood was not immediately available and the infant had to be given repeated exchange transfusions with incompatible group 0 Rh-negative blood. The infant is now healthy and thriving. The mother's red cells were subsequently found to lack all the antigens of the Rh system, and her serum reacted with all red cell samples except those of two unrelated Rhnull individuals. Her serum gave high titres (i.e. 1,024–4,096) by the indirect antiglobulin test against red cells of normal Rh phenotype, as well as against cells with partially deleted Rh phenotypes (titres = 128–512 with –D–·/–D– and ·D·/·D· samples, respectively), and was extremely active in antibody-dependent cell-mediated cytotoxicity and monocyte monolayer assays against red cells of normal Rh phenotypes.

We have recently detected a new qualitative variant of the RhE antigen characterised by negative reactions with a minority of anti-E sera, and an inability of the relevant cells to completely absorb most, if not all, anti-E samples that we tested. In tube tests, using a two-stage enzyme technique, cells from the proposita (V.R.) were completely agglutinated by 8 out of 12 undiluted polyclonal sera, weakly agglutinated (2+ and 3+ reactions) by 2 sera, but not agglutinated by the 2 remaining sera. Cross-absorption tests using 1 non-reacting, 1 weakly reacting and 4 strongly reacting polyclonal anti-E sera showed that V.R.'s cells only slightly reduced their anti-E titres versus r"r cells, whilst completely removing their reactivity against her own cells. Absorptions with R2R2 cells abolished activity against both cell types. The cells of V.R. failed to absorb or elute anti-Ew, and had normal expression of c. Three monoclonal human anti-E reagents were tested by fluorescence flow cytometry; one reagent failed to react with the cells of V.R., a second bound as strongly with V.R.'s cells as with R1R2 controls while the third bound much more weakly with V.R. than with any other E+ sample. The variant of RhE in the V.R. family is therefore distinct from Ew, and lacks an epitope recognised by a major constituent of most anti-E sera. This epitope is unlikely to be ET, since all of 91 samples from E-positive Australian aborigines (many of which are expected to lack ET) were agglutinated by a serum that failed to react with V.R.'s cells.(ABSTRACT TRUNCATED AT 250 WORDS)

A survey of sera containing antibodies to multiple low-incidence antigens revealed a variety of patterns of reactions with NFLD+, BOW+, and Wu+ red cell samples. Although NFLD, BOW, and Wu are distinct antigenic determinants (International Society of Blood Transfusion numbers 700.37, 700.46, and 700.13, respectively), their ability to absorb and elute "crossreactive" antibodies indicates a serologic and possibly a genetic relationship among the three.

The BCSH guidelines for FMH first published in 1999 are now being updated. Evidence on which some of the recommendations are made is being re-examined, e.g. the number of cells to be counted using the acid elution method (AE), to identify and quantify fetal bleeds. When using flow cytometry (FC) to measure FMH, the guidelines recommend that 500,000 events be collected, based on the statistics of rare event analysis where the CV% of the method decreases with the number of events collected (see Fig. 1). Extrapolating this statistical principle to acid elution (1 in 1000 ‘event’ equates to a 2.4ml bleed using Mollison's formula), where counting 6000 adult cells for FMH quantification is currently recommended, the CV% would be approximately 60 for a 4ml bleed. However this can be improved to approx. 25 if 10,000 cells are counted. (1) [ CV% (y-axis) vs number of events counted (x-axis) for a Poisson distribution where approx. 1 in 1000 positive events is expected. ] These statistics can also be applied to screening by FMH, where identification of bleeds >4ml is essential since the most commonly used post-natal dose of anti-D immunoglobulin (500 i.u.) covers a 4mL bleed. When scanning 25 low power fields (lpf) using x10 objective and x10 eyepiece, a minimum of 50,000 cells are screened (based on data from Mollison and UK NEQAS). The inter-laboratory CV% = 38 for bleeds of 2 - 6ml (by AE) in UK NEQAS exercises and the CV% according to rare event statistics is approximately 20. These figures can be used to predict the range of FMH values observed at different bleed volumes (see table), using the formula SD = (CV x mean) / 100 and where the expected bleed volume is the mean. mL FMH Expected range of FMH (mL) at 95% confidence (=2SD) CV% 20 (rare event) CV% = 38 (UK NEQAS) 4 2.4–5.6 1.0–7.0 2 1.2–2.8 0.5–3.5 Theoretically, 2ml is a safer trigger for quantification and follow-up, since with a 4ml ‘cut-off’ there is a risk that FMH of >4ml will not be recognised on screening 25 lpf, even where proficiency is good. Also, given the predicted high CV% for quantification by AE, the guidelines will recommend referral of bleeds >2ml for FC and instigation of follow-up, even where the standard anti-D dose exceeds 500 i.u.

Concordance of reactions of Wu+ and ‘Hov’+ cells with 153 sera containing multiple specificities to low-incidence antigens indicates that the ‘two’ antigens are identical. This conclusion is confirmed by absorption and elution studies.

TRALI (Transfusion Related Acute Lung Injury) occurs in 11–16 patients and is associated with 3–4 deaths per year in the UK. It is caused by HLA (e.g. anti-HLA-A2), neutrophil (e.g. anti-NA2) and leucocyte-panreactive (e.g. anti-5b) Ab. Recently an IvIg preparation enriched for HLA class I and II Ab caused TRALI in an untransfused male volunteer (Dooren et al. Lancet 1998; 358: 1601–1602). We therefore ascertained the incidence of class II Ab in sera in donors implicated in TRALI using commercial ELISA kits. Sera from 32 donors were tested for HLA class II antibodies using GTI B-Screen ELISA kits. All of these sera had previously been tested for leucocyte-reactive Ab using the GLAM assay, and for HLA class I Ab by LCT and ELISA (GTI QuikScreen® kits). The following reaction patterns were obtained: Hence HLA class II Ab were detected in 8/32 (25%) of sera implicated in TRALI. We suspect that they are more frequent than neutrophil-specific Ab in TRALI donors: we identified the latter in only 8/123 (6·5%) donors tested in the course of investigating 31 TRALI cases, i.e. in 6/21 (28%) donors who were LCT-GLAM+ and in 2/9 (22%) who were LCT+GLAM+.