Functional characteristics of red cells of sickle cell anemia patients with and without α-thalassemia-2

Oxygen equilibrium curves, red cell indices, 2,3-DPG levels, and the percentages of ISC and Hb F were determined for red cell fractions isolated by Dextran 40 density gradient centrifugation from blood of a sickle cell anemia (SS) patient, an SS patient with an additional homozygosity for α-thalassemia-2, and a normal control. In the SS patient, the MCHC and P₅₀ values increased and 2,3-DPG levels decreased with an increase in red cell density; the P₅₀ values were directly related to the MCHC values and inversely related to the 2,3-DPG levels. The bottom fraction contained about 80% ISC, and had the highest MCHC and P₅₀ values and the lowest 2,3-DPG level. In the SS α-thalassemia-2 patient, a decrease in the P₅₀ value corresponded with a decrease in 2,3-DPG levels because no change in MCHC value was observed. The bottom fraction contained about 10% ISC and had the lowest P₅₀ value. These data confirm that both MCHC and 2,3-DPG levels influence the oxygen affinity of SS red cells; an increase in MCHC decreases the oxygen affinity to a great extent while a concomitant decrease in 2,3-DPG level partially abolishes this effect.     

Oxygen dissociation curves in children with anemia and malignant disease

Two automatic apparatuses utilizing a dual wavelength spectrophotometer were used to perform oxygen dissociation curves on microsamples of blood. The method provides a complete print-out of an oxygen dissociation curve in 15--20 min and the P50 vlues obtained in normal individuals agree closely with those obtained by classical methods. These apparatuses were used to measure oxygen affinity, ie P50, in anemic children with malignant disease prior to treatment and in children undergoing therapy. Red cell 2,3-DPG levels were also measured. In patients with anemia at the time of diagnosis and prior to therapy, the P50 values and 2,3-DPG levels were elevated as is usual in other types of anemia. However, when oxygen affinity and 2,3-DPG levels were measured in anemic patients receiving treatment, three types of response to anemia were noted: 1) increased P50 and 2,3-DPG; 2) normal or low P50 and 2,3-DPG, and; 3) normal or low P50 with increased 2,3-DPG. Patients who adapted poorly to anemia during treatment had usually received prior intensive chemotherapy and transfusion therapy, and their pattern of red cell glycolytic intermediates was consistent with a red cell population with an increased cell age. The failure of some patients to respond to anemia with a decrease in oxygen affinity has implications in regards to the hemoglobin level at which they should be transfused.     

The Oxygen Affinity of Haemoglobin E

Oxygen dissociation studies were carried out on haemoglobin E (Hb E) at both high and low haemoglobin concentrations. Oxygen affinities of fresh red cells from three people homozygous for Hb E and from one with Hb E-beta thalassaemia (Hb-E trait/beta-thal trait) were low in three out of four patients studied, while the oxygen affinity of red cells from an individual with Hb-E was normal 2,3-DPG concentration in the fresh cells from the people with homozygous Hb E or Hb-E trait/beta-thal trait which showed low oxygen affinities were elevated sufficiently to account for the shifts observed. When the cells from two of these people with homozygous Hb E were depleted of 2,3-DPG. their oxygen affinities became the same as that of similarly treated normal cells. Pure 'stripped' Hb E in dilute solution behaved identically to Hb A in respect of P50, Bohr shift, haem-haem interaction, and interaction with inorganic phosphate or 2,3-DPG. Hb E, therefore, has the same oxypgen dissociation properties as Hb A both in dilute solution and in the red cell. The low oxygen affinities found in the fresh cells and in whole blood are caused by high 2,3-DPG concentrations within the cell.     

An adverse effect of insulin on the oxygen-release capacity of red blood cells in nonacidotic diabetics.

Oxyhemoglobin dissociation curves (ODC) were performed on blood from newly diagnosed, nonketotic diabetics prior to and following initial insulin treatment and from ambulatory juvenile diabetics before and after their usual morning insulin. In 10 newly discovered diabetics the average P50 at in vivo pH was normal prior to insulin (26.2 mm Hg), decreased to 24.5 mm Hg (p less than 0.005) on the day following the initial insulin administration, and was within normal limits (26.9 mm Hg) when the diabetes was finally well controlled and red cell 2,3-diphosphoglycerate (2,3-DPG) had risen to elevated levels. Oxygen affinity of hemoglobin was closely correlated with the content of red cell 2,3-DPG (r = 0.61, p less than 0.001) but was unrelated to the level of hemoglobin Alc. In 40 juvenile patients the average P50 was also normal prior to insulin administration but was significantly lower 3-4 hr after they had received their usual insulin dose (p less than 0.001). The study indicates that insulin administration to diabetics with high blood glucose levels may lead to transient decreases in red cell 2,3-DPG and in oxygen-releasing capacity of the red blood cells.     

In Vivo Aging of Transfused Erythrocytes and 2,3-Diphosphoglycerate Levels

Levels of 2,3-diphosphoglycerate (2,3-DPG)and oxygen affinity were measured in apatient with refractory congenital hypoplastic anemia in order to determine theresponse of aging transfused erythrocytesto a decrease in tissue oxygen tension.Ferrokinetic studies revealed that the rateof erythropoiesis after transfusion wasapproximately 10% of normal. In the interval between transfusions, the level of2,3-DPG and P₅₀ rose to values considerably above the normal range as the levelof the hematocrit decreased. On fiveseparate occasions, the 2,3-DPG concentration was elevated (average 25.5 µmoles/gof hemoglobin) and the P₅₀ level increased(average 33.2 mm of mercury) 24 hr priorto transfusion. It is concluded that in vivoaging of transfused erythrocytes does notprevent these cells from responding to anincreasingly hypoxic environment by elevating 2,3-DPG levels and by decreasingoxygen affinity.

Submitted on August 10, 1972 Revised on November 6, 1972 Accepted on November 18, 1972     

Red cell immunization in multiply transfused Malay thalassemic patients

The development of red blood cell (RBC) isoimmunization with alloantibodies and autoantibodies complicate transfusion therapy in multiply transfused thalassemia patients. Thus, the frequency, causes and prevention of these phenomena were studied among these patients. Clinical and serological data from 58 Malay multiply transfused thalassemic patients who sought treatment at Hospital University Sains Malaysia were collected and analyzed prospectively. Blood samples were subjected to standard blood bank procedures to screen for antibody and subsequent antibodies identification. All patients in our hospital received blood matched for only ABO and Rh (D) antigens. There were 46 (79.3%) patients with Hb E/beta thalassemia, 8 (13.8%) with beta thalassemia major, 3 (5.2%) with Hb H Constant Spring and 1 (1.7%) with Hb H disease. Overall, 8.6% of the patients had alloantibodies and 1.7% had autoantibodies. The alloantibodies identified were anti-E, anti-c, anti-K, anti-Jka, anti-N and anti-S. In conclusion, the transfusion of matched blood is essential for chronically multiply transfused patients in order to avoid alloimmunization. Considering the high frequency of anti E at our hospital, it is advisable to genotype patients and match the red cells for E antigens in multiply transfused thalassemia patients.     

Clinical and Laboratory Experience with Erythrocyte and Platelet Preparations from a 0.5CPD Erythro-Sol Opti System

Background: Red blood cells stored as concentrates or suspensions in additive solutions change rapidly their oxygen affinity mainly due to the loss of 2,3-diphosphoglycerate (2,3-DPG). When collected in CPD with half of the normal concentration of citrate and citric acid (0.5CPD) and stored in a new additive solution (Erythro-Sol), 2,3-DPG is better maintained. No studies of the oxygen affinity of red cells stored under these conditions have been published. In Erythro-Sol, red cells have a satisfactory in vivo recovery for 49 days but the conditions after 28 days, within which time most red cell units are transfused, have not been investigated. Of importance is also to be able to make platelet concentrates (PCs) from 0.5CPD blood. Little data are available concerning the clinical usefulness of platelets prepared from 0.5CPD buffy coats (BCs). Methods: Blood was collected in 0.5CPD, held at 20°C for 3–4 h, then separated with the bottom-and-top technique into red cells, plasma and BC. In a storage experiment with 6 U the 2,3-DPG and P₅₀ values were determined weekly and a number of in vitro parameters were tested on day 28. In 6 donors the in vivo recovery and survival of red cells were determined using a single-chromium technique. Transfusions of 212 0.5CPD-Erythro-Sol red cell units were given to hematological patients under supervision. PCs derived from pools of 0.5CPD BCs suspended in PAS2 (T-Sol) were transfused to 20 thrombocytopenic patients and compared with CPD-BC-PCs suspended in PAS1. Corrected count increments (CCI) were determined. Results: The erythrocyte 2,3-DPG and P₅₀ values were normal or slightly subnormal initially but increased to supernormal levels during the 1 week, and remained at these levels for a further 1–3 weeks; the 2,3-DPG was two thirds of normal after 28 days, the P₅₀ was 3.72±0.28 kPa after 14 days and 2.84±0.41 after 28 days (mean ± SD). The P₅₀ values corresponded closely (r²= 0.903) to 2,3-DPG. The in vivo recovery of 4-week-stored red cells was 89.6±5.5% and the T₅₀ was 32.2±2.0 days. No adverse effects were observed in the transfusions. The CCI values did not differ between test and control groups; in both, 3- to 5-day-stored PCs gave lower CCI than fresh (0–2 days) PCs. Patients with acute myeloid leukemia AML (n = 11) had significantly lower CCI values than patients with myelodysplastic syndrome, myeloma and lymphoma (n = 9; CCI_{1 h}: p = 0.001; CCI_{24 h}: p = 0.006). Conclusions: Red cells stored in Erythro-Sol sustain a normal or slightly lowered oxygen affinity for 2–4 weeks, their viability is excellent, and they are well tolerated in clinical transfusions. Platelets prepared from 0.5CPD-BCs cause CCl_s of the same magnitude as CPD-BCs.     

Erythroid marrow activity and functional anemia in patients with the rare interaction of a single functional a-globin and beta-globin gene

BACKGROUND AND OBJECTIVES: The degree of globin chain imbalance and tissue hypoxia are important determinants of clinical severity in thalassemia syndromes. Thus phenotypic expression may be modified by interaction of alpha- and beta-thalassemia defects, level and type of hemoglobin synthesized and oxygen release to the tissues. We evaluated hematology, erythroid marrow activity and functional anemia in patients with the rare interaction of a single a-globin gene and heterozygous beta-thalassemia (HbH/beta-thal trait). DESIGN AND METHODS: In 7 patients characterized by DNA analysis to have HbH disease genotypes with beta-thalassemia trait, we assessed hematologic findings, serum transferrin receptor (sTfR), serum erythropoietin (Epo), red cell 2,3-disphosphoglycerate (2,3-DPG) and whole blood oxygen releasing capability. RESULTS: Patients with HbH/beta-thal trait had moderate anemia, marked hypochromasia and microcytosis, normal or raised HbA2, and no electrophoretically/chromatographically detectable HbH. Epo and sTfR levels were significantly higher than in beta-thalassemia heterozygotes, but lower than in patients with HbH disease; 2,3-DPG levels were highest in HbH/beta-thal trait. Oxygen binding studies and simulations showed reduced oxygen affinity (P50) in HbH/beta-thal trait, resulting in increased oxygen release (O2R). INTERPRETATION AND CONCLUSIONS: Hematologic findings and bone marrow activity in patients with HbH/b-thal trait were consistent with the modified globin chain imbalance and hemoglobin synthesis expected from interaction of HbH disease with heterozygous b-thalassemia, although this rare complex genotype may elude diagnosis based on hematology alone. Significantly higher red cell 2,3-DPG levels were an unexpected finding, and the consequent increase in oxygen release capability resulted in a compensated functional anemia relative to hemoglobin levels.     

The oxygen transport system of red blood cells during diabetic ketoacidosis and recovery

Daily evaluations of 8 newly detected ketoacidotic diabetics showed the Bohr-effect of haemoglobin to be decreased by 50% while erythrocyte 2,3-DPG was decreased below 10 mumoles/g Hb. 2,3-DPG correlated strongly with pH during acidosis and with plasma inorganic phosphate (Pi) subsequently to the first insulin administration. Oxygen affinity of haemoglobin, measured as P50 act pH, was unchanged in ketoacidosis compared to the time, however, P50 act pH fell striking (p less than 0.001) and remained decreased up to 7 days depending upon the resynthesis of 2,3-DPG in relation to Pi. The Hill-coefeficient in reflecting the slope of the oxygen dissociation curve was diminished in ketoacidosis (p less than 0.005), and decreased further after pH-normalization (p less than 0.005). There was a close association of n with 2,3-DPG (p less than 0.001) and additionally with Pi at 2,3-DPG-levels below 10 mumoles/g Hb. Based on these findings a decreased erythrocyte oxygen release of one fifth during acidosis and more than one third after pH-correction can be hypothesised. In view of the intimate relation of Pi to the oxygen transport system it is suggesed that treatment of ketoacidosis should include Pi-sugstitution.     

Oxygen Binding to Haemoglobin in Subjects with Hypoproliferative Anaemia, with and without Chronic Renal Disease: Role of pH

S ummary . The relationships of haenioglobin concentration and blood pH to red cell 2,3-dipliosphoglycerate and oxygen binding by haenioglobin have been studicd in healthy subjects and subjects with hypoproliferative anaemia with or without severe chronic renal disease. Red cell 2,3-DPG was inversely correlated with haemoglobin deficit and directly and equally strongly associated with blood pH in anaemic subjects without chronic renal disease. In subjects with chronic renal disease receiving regular haemodialysis, predialysis pH was not increased despite severe anaemia, and red cell 2,3-DPG was not significantly elevated, except in subjects who had a sustained alkalosis due to the use of sodium bicarbonate.
In hypoproliferative anaemia, the incrcmcnt in pH was associated with the decrease in haenioglobin concentration such that 80% of the increase in p50 measured at standard conditions which occurred with anaemia was explicable by the relationship of (a) pH with haemoglobin concentration, (b) red cell 2,3-DPG with pH, and (c) pso std with red cell 2,3-DPG. However, p50 at the pH and base excess present in vivo was similar in all anaemic subjects whether an increase in red cell 2,3-DPG occurred or not. Blood alkalosis and the accumulation of 2,3-DPG cancelled each other's effect on oxygen binding by haemoglobin. Hence, increased red cell 2,3-DPG and p50 compensated for the alkalosis of hypoproliferative anaemia, not for the deficit in haemoglobin concentration.     

Red Cell 2,3-Diphosphoglycerate Levels in Children with Hereditary Haemolytic Anaemias

The role of red cell 2,3-diphosphoglycerate (2,3-DPG) in increasing the availability of haemoglobin oxygen in neonatal jaundice and hereditary haemolytic anaemias was investigated. Measurements of 2,3-DPG were carried out on 58 normal children and six normal adults, 18 full-term newborns with neonatal jaundice and 57 cases (51 children and six adults) with hereditary haemolytic anaemias. In normal children and adults, with a mean haemoglobin of 12.69 g/dl, mean 2,3-DPG was 14.90 mumol/g Hb. In jaundiced newborns with a mean haemoglobin of 16.04 g/dl mean 2,3-DPG levels were 14.51 mumol/g Hb, i.e. normal. 2,3-DPG levels were increased in patients with beta-thalassaemia major, alpha-thalassaemia, sickle-cell disease, favism, hereditary spherocytosis and in heterozygotes for beta-thalassaemia with increased haemoglobin F. In heterozygotes for beta-thalassaemia with increased haemoglobin A2 only and in sickle cell trait 2,3-DPG levels were normal.     

Phosphoglycolate phosphatase and 2,3-diphosphoglycerate in red cells of normal and anemic subjects

Red cell phosphoglycolate phosphatase (PGP) and 2,3-diphosphoglycerate (2,3-DPG) were investigated in normal and anemic patients and rabbits. In hemolytic anemia and blood-loss anemia, characterized by a young red cell population, there was an increase in both phosphoglycolate phosphatase activity and 2,3-diphosphoglycerate levels. In aplastic anemia, the phosphoglycolate phosphatase activity was normal, but the 2,3-diphosphoglycerate values were nonetheless increased. Thus, no relationship was found between phosphoglycolate phosphatase activity and 2,3-diphosphoglycerate levels. The lack of correlation between the activity of phosphoglycolate phosphatase and 2,3-DPG levels suggests that modulation of phosphoglycolate phosphatase activity does not control the level of 2,3-DPG in erythrocytes.     

Oxyhemoglobin dissociation curve and 2,3-diphosphoglycerate in chronic hypoxemia

The measurement of the oxyhemoglobin dissociation curve (ODC) and 2,3-diphosphoglycerate (2,3-DPG) in patients with chronic hypoxemia is important from the view point of tissue oxygenation. However, there have been no consistent results that explain the relation among chronic hypoxemia, 2,3-DPG and P50, which is oxygen pressure at an oxygen saturation of 50 percent. The aim of this study is to clarify what factors affect P50 and 2,3-DPG. 1) Patients with chronic hypoxemia, who showed PaO2 less than 60 Torr, had significantly higher P50 than normal subjects. 2) The concentration of Hb showed significant negative correlation with both P50 and 2,3-DPG. 3) Arterial blood pH showed significant positive correlation with both P50 and 2,3-DPG. 4) In a group with normal levels of Hb and pH, there was significant negative relationship between PaO2 and P50. 5) In a group with normal levels of Hb and pH, there was significant positive relationship between PaCO2 and P50. 6) In a group with normal levels of Hb, pH and PaCO2, there was significant negative relationship between PaO2 and 2,3-DPG. In conclusion, P50 and 2,3-DPG are affected largely by Hb concentration or blood pH, with or without hypoxemia. However there is a mechanism by which P50 and 2,3-DPG are increased by hypoxemia itself in a group with normal levels of Hb, pH and PaCO2.     

Serum transferrin receptor as a marker of erythropoiesis suppression in patients on chronic transfusion

In the management of patients requiring chronic transfusion, various parameters may be used to evaluate the degree of erythroid marrow suppression. The aim of our study was to assess which of these parameters provide the most useful assessment of erythropoiesis. We studied 27 chronically transfused patients, 19 with sickle cell disease (SS patients) and 8 with thalassemia. Thirty-one nonchronically transfused SS patients and 74 healthy children served as controls. We measured serum transferrin receptor levels, reticulocyte counts, hemoglobin (Hb) concentrations and erythropoietin levels. The serum transferrin receptor levels were very elevated in control SS patients and remained significantly elevated in those on transfusion therapy, but were normal in thalassemia patients, indicating a more complete suppression of erythropoiesis. The reticulocyte counts were elevated in all SS patients, even when on chronic transfusion, but were in the normal range in patients with thalassemia. Erythropoietin levels were elevated in patients with thalassemia and in all the SS patients. Hb levels negatively correlated with serum transferrin receptor and erythropoietin in all SS patients. In the transfused SS patients, a higher HbS level correlated with higher reticulocyte counts, transferrin receptor, and erythropoietin levels. In thalassemia patients, erythropoiesis was more completely suppressed, as reflected both by normal reticulocyte counts and near-normal transferrin receptor levels. Though the reticulocyte counts were not significantly different in the transfused SS patients, the serum transferrin receptor levels were less elevated than in SS patients not on transfusion. The serum transferrin receptor level appears to be the most useful marker of marrow erythropoietic activity in chronically transfused SS patients. We recommend that reticulocyte counts be integrated with periodic measurements of serum transferrin receptor levels. Am. J. Hematol. 60:121–125, 1999. © 1999 Wiley-Liss, Inc.     

The Oxygen Affinity of Haemoglobin Hammersmith

Oxygen dissociation studies were carried out on red cells and lysates from a patient heterozygous for Hb Hammersmith. The oxygen affinity of the cells at pH 7.1 was decreased, partly by an increased cellular concentration of 2,3-DPG and more importantly by an intrinsic low affinity of Hb Hammersmity. Haem-haem interactions were reduced and an abnormal Bohr effect (pH 7.1-7.4) was found in the cells which would bring about an additional decrease in oxygen affinity at physiological pH. Oxygen dissociation studies on the lysates showed a low oxygen affinity, a normal Bohr effect at 50% saturation, slightly decreased haem-haem interactions and a normal interaction with 2,3-DPG. Estimation of the percentage of Hb Hammersmith in the lysates of the patient's cells by selective precipitation of the abnormal beta chain with p-chloromercuribenzoate (PCMB) showed the presence of 30-33% abnormal haemoglobin. Assuming no interaction between the Hb A and the Hb Hammersmith, oxygen dissociation curves were calculated for Hb Hammersmith alone and these indicated that it has greatly decreased oxygen affinity, a normal Bohr effect, normal 2,3-DPG interaction and (somewhat) decreased haem-haem interactions.     

Anemia in Patients with Coinherited Thalassemia and Glucose-6-Phosphate Dehydrogenase Deficiency

Thalassemia and glucose-6-phosphate dehydrogenase (G-6-PD) deficiency are genetic disorders that cause hemolytic anemia. In areas with high frequencies of both hematological disorders, coinheritance of G-6-PD deficiency with thalassemia can be found. Whether G-6-PD deficiency, coinherited with thalassemia, enhances severe anemia is still unclear. Hematological parameters between thalassemia carriers with G-6-PD deficiency and those without G-6-PD deficiency were compared. The G-6-PD deficiency was diagnosed in 410 blood samples from thalassemia patients using a fluorescent spot test. The levels of hemoglobin (Hb), packed cell volume (PCV), mean corpuscular volume (MCV) and Hb A₂/Hb E [β26(B8)Glu→Lys; HBB: c.79G>A] were measured using an automated blood counter and high performance liquid chromatography (HPLC), respectively. The G-6-PD deficiency was found in 37 samples (9.02%). Mean levels of Hb, PCV, MCV and Hb A₂/E were similar between the two groups. Thus, G-6-PD deficiency did not enhance red blood cell pathology or induce more anemic severity in thalassemia patients.     

Red Cell ATP and 2,3-Diphosphoglycerate Concentrations as a Function of Dihydroxyacetone Supplementation of CPD Adenine

Abstract. Units of CPDA-1 whole blood were subdivided and each treated with additions of dihydroxyacetone (DHA) to give final concentrations from 0 to 80m M . The 'optimum' concentration of DHA to maintain 2,3-diphosphoglycerate (2,3-DPG) with minimal loss of ATP during 42 days of storage appeared to be 30m M of DHA. With this formulation, red cell 2, 3-DPG concentrations rose to 130–140% of normal by 14 days and then decreased in a near-linear manner to 50–60% normal by 42 days, while maintaining adequate ATP levels. In addition, packed red cells were prepared from CPD fresh blood and treated with adenine, glucose, and various concentrations (0-80m M of DHA. The cells also responded most favorably to 30m M DHA, although the response was not as positive as whole blood. This concentration of DHA produced nearly 100% maintenance of 2,3-DPG at 14 days with subsequent fall to 30% of normal by 42 days.     

Non-enzymatic glycosylation influences Hb S polymerization

In vivo glycosylated components of Hb S were isolated from red cell hemolysates of sickle cell anemia patients by application of affinity chromatographic and cation exchange chromatographic techniques. The total glycosylated fraction (GHb) of the whole hemolysate, Hb SIc, Hb So-glycosylated (formed mostly by glycosylation epsilon-NH2 groups of lysyl residues), and Hb So-nonglycosylated fractions were isolated in this manner. GHb contained about 33% Hb SIc and 42% Hb So (Hb So-glycosylated) and the rest was glycosylated Hb F and Hb A2. As expected, the binding of 2,3-DPG was affected only in Hb SIc and not in Hb So-glycosylated. Hb SIc and Hb So-glycosylated had higher solubility in concentrated phosphate solutions and had higher minimum gelling concentrations than the non-glycosylated form of Hb So. These effects are interpreted to be due to modification by glycosylation of specific sites that are directly or indirectly involved in the intermolecular contacts.     

HLA class I antibodies in thalassemic patients

This study was undertaken to demonstrate the prevalence of HLA class I antibodies among 62 polytransfused patients. The diagnosis included beta-thalassemia major, beta-thalassemia/Hb E disease and severe Hb H disease. Their ages ranged from 1 year to 23 years with the mean age of 10.7 years. The number of packed red cell transfusions ranged from 3 to 235 with the mean of 60 episodes per patient. The standard microlymphocytotoxicity test was performed using 50 panels of lymphocytes which specifically identified the majority of HLA class I antibodies. 31/62 cases (50%) were positive for HLA class I antibodies. The detection of single or multiple antibodies depended upon the number of blood transfusions and the patients' ages. These antibodies were induced by the leukocytes present in the transfused packed red cells. Therefore, leukocyte-reduced packed red cells prepared by either additional inverted centrifrugation or leukocyte filter is suggested for the routine blood bank service.     

Effectiveness of Red Blood Cells Filtered Through Cotton Wool to Prevent Antileukocyte Antibody Production in Multitransfused Patients

Abstract. The effectiveness of red blood cells made leukocyte-free by filtration through cotton wool to prevent the production of antileukocyte antibodies was evaluated in children suffering from Cooley's anemia. Two studies were performed: study I was carried out prospectively in two groups of non transfused patients, one group treated with leukocyte-free filtered red cells, the other with buffy-coat-free packed red cell units. Different types of antileukocyte antibodies were looked for in both groups and the results were compared. In study II the behavior of pre-existing lymphocytotoxic antibodies found in the serum of children previously transfused with standard or buffy-coat-free packed red cell units was followed after the patients had been passed to a program of transfusion with leukocyte-free filtered red cells. Study I showed that none of the patients transfused with leukocyte-free filtered red cell units have produced antileukocyte antibodies, while these could be found in 2/3 of the patients transfused with buffy-coat-free packed red cell units. Study II showed that the repeated transfusion of leukocyte-free filtered red cells to patients who possessed in their serum preformed lymphocytotoxic antibodies did not cause any increase in the potency or spectrum of these antibodies, but was in fact accompanied in some cases by their decrease or disappearance. It is concluded that filtration through cotton wool is an easy and inexpensive means of preparing leukocyte-free red blood cells for transfusion capable of preventing (or reducing) the production of antileukocyte antibodies in multitransfused patients.