异体骨髓移植供者骨髓采集方法和安全性探讨

目的：评价异体骨髓移植供者骨髓采集方法的安全性。方法：在连续续硬外或全身麻醉下对408例供者进行骨髓采集术，术中回输预先采集的供者自体血。结果：术中抽取骨髓血的中位数为1108ml，输血输液量的中位数为4000ml，骨髓采集术中术后不良反应主要有低血压，骨痛和发热，发生率分别为2．5％，100％及26．5％，全部供者均未见严重并发症，结论：连续硬膜外麻醉下进行骨髓采集术安全可靠，并发症发生率较少，术中采用预贮式自体输血有很大优势。     

ABO-incompatible pediatric bone marrow transplantation: a simple method to remove red blood cells from small volume marrow grafts.

A simple and reliable technique for removal of ABO incompatible marrow red cells is described. This method requires a blood cell processor and third party red cells are used as a shelf. Using this technique, nine children age 0.5-11.5 years received allogeneic bone marrow transplantation from ABO incompatible donors. A median of 4.8 ml of incompatible red cells were transfused. There was no evidence of a hemolytic transfusion reaction in any patient. A median of 75% of nucleated marrow cells were recovered and used for transplantation. Engraftment occurred at the same time as with ABO compatible transplants. Autologous marrow red cells were reinfused into four young donors. This 'shelf' technique for red cell depletion is an acceptable method for processing small volume, ABO incompatible marrow harvests from pediatric donors.     

Collection of autologous blood for bone marrow donation: how useful is it?

The hospital charts of 495 adult bone marrow (BM) donors to adult patients were reviewed to determine how necessary it is to collect autologous blood for marrow donation. An autologous transfusion was given to 79% of the donors. The median total volume of marrow harvested was 900 ml (range 450-1350 ml). The median number of nucleated cells harvested was 3.2 x 10(8)/kg patient weight (range 0.9-7.4 x 10(8)/kg patient weight). On the morning following the harvest, the median haemoglobin (Hb) concentrations were 104 g/l (79-135 g/l) in the female and 122 g/l (89-151 g/l) in the male donors autotransfused, and 96 g/l (75-127 g/l) in the female and 119 g/l (88-141 g/l) in the male donors not autotransfused. The post-donation Hb was lower than 85 g/l in four and lower than 90 g/l in 25 donors. Of the 25 donors with post-harvest Hb lower than 90 g/l, 23 were females and 14 had received an autologous transfusion. This study shows that, with a few exceptions, it is not necessary to collect autologous blood from healthy BM donors before the marrow harvest. The post-donation Hb concentrations do not decrease to levels detrimental to healthy persons whether autologous blood is transfused or not.     

Transfusion of bone marrow red cells during bone marrow harvests

Large volumes of bone marrow may be required for certain types of autologous bone marrow transplants. The present study was done to determine whether red cells obtained during a bone marrow harvest would be useful in reducing homologous transfusion requirements. A group of patients receiving standard transfusion support during the harvest (group 1) was compared to a group that received processed bone marrow red cells (PBMRBC) (group 2). Using the Cobe 2991 cell processor, 90% of the harvested bone marrow red cells were extracted and transfused during the procedure. Group 2 received a median of 1500 ml of blood processed from the bone marrow or 413 ml (volume of marrow processed x hematocrit) of red cells. Infusion of the PBMRBC reduced the homologous transfusion requirement from 6.5 units to 3.0 units (p = 0.02). In addition, group 1 had a 20% decrease in hematocrit following transfusion compared to the pre-harvest hematocrit, as opposed to an 8% decrease in group 2 (p = 0.02). This study indicates that PBMRBC can reduce the homologous transfusion requirements during an autologous bone marrow harvest.     

Experiences of the first 493 unrelated marrow donors in the National Marrow Donor Program

More than 410,000 people participated in the National Marrow Donor Program (NMDP) as of October 1, 1991, and more than 850 volunteers had donated marrow. While the incidence of serious morbidity as a result of bone marrow donation is rare, the incidence of lesser complications and the long-term consequences of marrow donation are not known. To determine the incidence of donor complications and measure the recovery time of volunteer, unrelated marrow donors, we analyzed the results of surveys of the first 493 persons who donated marrow through the NMDP. The marrows were collected at 42 centers. The median age of the donors was 37.9 years (range 19.1 to 55.6 years). The median volume of marrow collected was 1,050 mL (range 180 to 2,983 mL). Autologous red blood cells were transfused to 89.8% (439) of donors but only 0.6% (3) of donors received allogeneic blood. Acute complications related to the collection procedure occurred in 5.9% of donors; but a serious complication, apnea during anesthesia, occurred in only one donor. When donors were questioned approximately 2 days following discharge from their hospitalization, most donors described symptoms related to the collection; 74.8% experienced tiredness, 67.8% experienced pain at the marrow collection site, and 51.6% of the donors experienced low back pain. Donors were surveyed repeatedly until they felt that they had recovered completely. Mean recovery time was 15.8 days; however, 42 (10%) donors felt that it took them > or = 30 days to recover fully. The duration of the marrow collection procedure and duration of anesthesia both positively correlated with donor pain and/or fatigue following the collection; but the duration of the collection procedure had the highest correlation with post-collection pain and fatigue. The volume of marrow collected per unit of donor weight was more weakly correlated with donor pain and/or fatigue than the anesthesia and collection times. When multivariate analysis was used to analyze the correlation between donor recovery time and these variables, only the duration of the collection was found to correlate significantly with donor recovery time (P = .001). This analysis demonstrates that marrow donation is well tolerated with few complications. To decrease further the incidence of donor discomfort and recovery time following donation, the duration of the collection procedure, and probably the duration of anesthesia, and the volume of marrow collected, should be kept to a minimum.     

Acceleration of hemopoietic recovery after autologous bone marrow transplantation by low doses of peripheral blood stem cells.

Twenty patients with advanced malignant disease submitted to autologous bone marrow transplantation with marrow either unpurged (10 patients) or purged in vitro with mafosfamide (10 patients) after ablative chemotherapy, received simultaneously autologous peripheral blood stem cells (PBSC) collected during one to three 3 h cytapheresis procedures. The kinetics of the hematological recovery of these patients were compared to those of a group of patients suffering from similar diseases and grafted in the same institution with either unpurged marrow only (14 patients) or purged in vitro with mafosfamide (six patients). The median times to reach 10(9)/l leukocytes, 0.5 x 10(9)/l polymorphs, and 50 x 10(9)/l platelets were reduced by 10, 10, and 13 days, respectively, in patients transfused with both autologous bone marrow and peripheral blood stem cells as compared to those receiving bone marrow only. A reduction in the numbers of days spent in hospital post-transplantation (p less than 0.01), of days of fever greater than 38 degrees C (p = NS), and of platelet (p = 0.07) and of red blood cell transfusions (p less than 0.01) were also observed in the group of patients grafted with bone marrow and PBSC.     

Alloimmunization to RhD by Platelet Transfusions in Autologous Bone Marrow Transplant Recipients

Platelet transfusions from RhD-positive (D-positive) donors are often given to RhD-negative (D-negative) cancer patients. The low observed rate of alloimmunization has been attributed to disease and therapy-related immunosuppression. We have studied the occurrence of alloimmunization in 16 D-negative patients who did not have detectable anti-D prior to autologous bone marrow transplantation for malignant disease. All received D-positive platelets, but no other D-positive blood product. Three patients (19%) developed anti-D at 13, 24 and 83 days, respectively, after first receiving D-positive platelets, and after a total dose of 53, 65 and 119 D-positive platelet unit equivalents, respectively. Two of them also developed anti-C. The 13 patients in whom anti-D was not detected were also heavily transfused with D-positive platelets (mean +/- SD = 136 +/- 82 platelet unit equivalents). In 6 of them, the last recorded antibody screen was less than 3 months after the first D-positive platelets, and may not exclude a primary immune response. Thus, despite profound immunosuppression associated with autologous marrow transplantation, alloimmune responses to D-positive red cells in platelet concentrates can occur in some D-negative recipients.     

Cytotoxic T-lymphocyte response to cytomegalovirus after human allogeneic bone marrow transplantation: pattern of recovery and correlation with cytomegalovirus infection and disease.

The high rate of severe cytomegalovirus (CMV) disease after bone marrow transplantation (BMT) is related to the profound immunodeficiency posttransplant. Because cytotoxic T lymphocytes (CTL) have been implicated in resistance to viral infections, we examined the restoration of the CMV-specific CTL response in 20 patients who received bone marrow from HLA-matched, CMV-seropositive donors. Blood specimens were obtained from patients at 1, 2, and 3 months after BMT and from the donors on a single occasion. Peripheral blood mononuclear cells were cocultured with CMV-infected donor-derived fibroblasts for 2 weeks and then tested for cytotoxicity against CMV-infected and uninfected autologous or HLA-mismatched fibroblasts. Cytolytic activity was detected in all 20 donors, with specificity for autologous CMV-infected targets demonstrable in 17 (median CMV-specific lysis at an effector:target ratio of 15:1 was 32%, range 18% to 83%). Specific lysis was mediated by CD8+, class I-restricted T cells, as shown by inhibition with anti-class I monoclonal antibody and by selective depletion of effector cells. By contrast, CMV-specific CTL were detected in only 10 of 20 patients after BMT (median lysis 29% at 3 months post-BMT). None of these 10 patients developed CMV pneumonia, whereas 6 of the 10 patients with an undetectable CMV-specific CTL response after BMT died with CMV pneumonia. These results demonstrate that restoration of CMV-specific CTL responses may require an extended time period after BMT in some patients, and that such patients are at increased risk of developing severe CMV disease. Approaches to reconstitute CMV immunity in BMT patients by adoptive transfer of CMV-specific CD8+ CTL clones derived from the bone marrow donor are now being pursued.     

Unrelated allogeneic bone marrow donation: short- and long-term follow-up of 103 consecutive volunteer donors

Between 1992 and 1999, 105 unrelated allogeneic bone marrow collections from 103 volunteer donors (65 males and 38 females; median age 33 years) were carried out in three northern Italian centers (Verona, Bolzano and Padova) affiliated with the Italian Bone Marrow Donor Registry (IBMDR). The average volume of BM collected was equivalent in both genders (1143.1 ml for males and 1054.2 ml for females; P = 0.1), although the average volume collected for unit of body weight and the average post-collection blood volume depletion was higher in females (respectively 17.1 ml/kg and 14.2% in females, 14.8 ml/kg and 12% in males; P= 0.01 and 0.03). There was no statistically significant difference between males and females in the total number of nucleated cells collected. We did not record any acute life-threatening event during or after the bone marrow collections. The most frequent complaint was pain at the collection site (77%) followed by the onset of fatigue (38%) and nausea and vomiting (25%); all of these were short-term problems. Hospitalization was short (average 20.2 h) and donors started their normal daily activities after an average of 5.4 days. We also monitored Hb, serum ferritin levels, WBC and platelet counts in the post-collection period (average follow-up 40.1 months). All donors signed a written informed consent for a further bone marrow collection, if needed. Our findings confirm the short- and long-term safety of allogeneic bone marrow collection in volunteer donors.     

Transplantation of major ABO-incompatible bone marrow depleted of red cells by hydroxyethyl starch

23 bone marrow transplant recipients whose donors where major ABO-incompatible received marrow depleted of red cells prior to infusion utilizing gravity sedimentation in hydroxyethyl starch. The in vitro red cell-depletion procedure effectively removed 97.8% (mean) of the red cells from the harvested marrows and preserved 86.8% of the nucleated cells and 98.2% of the CFU-C activity in 25.4% of the original volume. All recipients had a significant quantity of isohemagglutinins of both IgM and IgG classes demonstrable in their serum at the time of the marrow infusion. Patients were premedicated and well-hydrated prior to the infusion and tolerated the infusion well. These patients demonstrated bone marrow engraftment at the same rate as did those patients whose marrow donors were either ABO-identical or minor ABO-incompatible. There was no difference in the incidence of or time to development of graft versus host disease, the incidence of graft rejection, or patient survival among the groups. Recipients of red cell-depleted major ABO-incompatible bone marrow transplants demonstrated production of donor-type red cells somewhat later and required slightly more red cell transfusion support that did the other groups of recipients. This red cell-depletion technique is safe and effective in the management of major ABO-incompatible bone marrow transplantation.     

CFU-C Populations in Blood and Bone Marrow of Dogs after Lethal Irradiation and Allogeneic Transfusion with Cryopreserved Blood Mononuclear Cells

Colony forming units in agar (CFU-C) were assayed in both bone marrow and peripheral blood of dogs during haemopoietic recovery after lethal total-body irradiation (1200 R) and allogeneic transfusion of blood mononuclear cells (MNC) from histocompatible donors. MNC had been collected from the peripheral blood by continuous-flow centrifugation leucapheresis and cryopreserved at -196°C until transfusion. Two groups of dogs were studied. Group 1 dogs (n = 12) were given between 0.39 and 2.76 times 10⁹ MNC per kg body wt. Group 2 dogs (n = 14) were transfused with a similar number of MNC, ranging from 0.51 to 1.87 times 10⁹ per kg body wt., but in addition underwent immuno-suppressive therapy with methotrexate. In group 1 dogs, there was a rather good correlation between the number of CFU-C in the regenerating bone marrow and the recovery of the peripheral blood granulocyte values. The regeneration of the CFU-C population in the bone marrow of methotrexate-treated dogs showed a somewhat more heterogeneous picture than in dogs of group 1 and in dogs that, in a previous study, were transfused with autologous MNC. The minimum time interval required for the reconstitution of peripheral blood CFU-C to normal levels was 2–4 weeks but usually took from 4–14 weeks.     

Bone Marrow Fractionation by the Haemonetics System: Reduction of Red Cell Mass before Marrow Freezing, with Special Reference to Pediatric Marrow Volumes

For purposes of freezing autologous marrow or transplants of allogeneic marrow with major ABO blood group incompatibility, 54 freshly harvested bone marrows from children of 7-65 kg of weight were depleted of their red cells with the Haemonetics V50 system. The marrow volumes ranged from 230 to 1,145 ml, with 17 small (200-399 ml), 18 intermediate (400-799 ml) and 19 large (800-1,200 ml) volumes. After processing, the median recoveries were: volume 24%, red cell mass 18%, and nucleated cells 75%. In the small marrow volume group, a good nucleated cell recovery was achieved at the expense of red cell depletion. The colony-forming units, granulocytes-macrophages (CFU-GM) were normal after thawing of processed, cryopreserved marrows, and good engraftment of both allogeneic and autologous marrows were achieved. We conclude that marrow processing with the Haemonetics V50 system results in adequate red cell depletion and good nucleated cell recovery without open-air contact of marrow or excessive time consumption. For small marrow volumes, however, the red cell depletion was suboptimal, and a bowl size smaller than 125 ml is desirable for pediatric use.     

Harvesting marrow for autologous transplantation from patients with malignancies

One hundred and sixty-six patients with malignancies had 170 consecutive bone marrow collection procedures in anticipation of autologous transplantation. The morbidity associated with these harvest procedures was increased compared to morbidity experienced by normal donors who had marrow collected for allogeneic transplantation when unexplained fever (p = 0.02) and infection (p = 0.008) were considered. No differences could be found in the characteristics of the harvested marrow including concentration of nucleated cells and nucleated cell content between the two donor groups. No deaths occurred as a result of the harvest procedure. However, a previously reported series of autologous marrow donors from a different institution did not demonstrate an increased morbidity associated with marrow harvesting when compared to the same series of normal donors. The differences found in the two autologous series might be explained by differing characteristics of the patient populations. The increased morbidity found in our patients was tolerable and reversible and not a contraindication to high-dose therapy with autologous bone marrow transplantation.     

Prevention of primary transfusion-associated cytomegalovirus infection in bone marrow transplant recipients by the removal of white cells from blood components with high-affinity filters

Summary. A prospective study was carried out to determine whether use of cytomegalovirus (CMV) unscreened red blood cells and platelet concentrates, white blood cell (WBC) depleted with high-efficiency filters, would prevent transfusion-associated (TA) CMV infection in CMV seronegative bone marrow transplant recipients. Blood components were filtered in the bloodcentre under quality control and after filtration residual WBC counts were always below 5 × 10⁶ cells/U. Since 1990, 23 consecutive allogeneic and 37 autologous CMV seronegative marrow transplant recipients, have been transfused with filtered blood components and followed for 6 months for evidence of CMV infection by monitoring culture and CMV serology. None of the patients showed clinical symptoms of CMV infection, and CMV cultures during episodes of fever were always negative. IgM anti-CMV antibodies were negative during the study in all patients. Low titres of IgG anti-CMV antibodies (5-12 relative ELISA units) were found in 24/60 patients during the first month after bone marrow transplantation (BMT), probably due to passive transfer of IgG administered with the platelet transfusions. 3 and 6 months after BMT, 56 and 48 patients respectively were still alive; and CMV serology was negative in all patients. The results show that TA-CMV infection is preventable by filtration of blood through high-efficiency filters in patients undergoing autologous and allogeneic BMT.     

自体混合HLA半相合异基因骨髓移植后的移植物抗宿主病

作者在动物实验的基础上,成功地用自体骨髓混合HLA半相合异基因骨髓移植治疗了16例恶性血液病患者。本文报告这些患者混合移植后GVHD的发生情况及其对疗效的影响。结果证实移植过程是安全的,16例中无1例发生aGVHD。但在ABMT后2h内输入异基因骨髓的8例皆有不同程度的cGVHD,主要表现在皮肤、粘膜炎,全血细胞减少,肝功异常,发热、体重减轻及易感冒等,中位随访13.5个月无1例复发,除2例在移植后3和8月分别因爆发肝炎和急性阑尾炎穿孔死亡外,余6例皆存活,最长1例已无病存活19月余,其中供受者性别不同的6例移植后性染色体观察3例皆形成嵌合体,最长者12月余。而ABMT移植后6h输异基因骨髓的8例,未观察到急慢性GVHD,其中5例4～7月后复发,仅1例仍持续缓解,其中5例染色体检查未形成嵌合体,提示ABMT后2h内输入异基因骨髓的混合移植,可诱发早发的cGVHD,部分受体内可形成嵌合体,减少白血病的复发,并发症较少而较轻,可望成为恶性血液病治疗的新的更有效的途径。     

Relationship between differing volumes of bone marrow aspirates and their cellular composition

We compared the cellular composition of the first 1.0 ml volume bone marrow aspirate with that of an aliquot from the total bone marrow harvest at the end of the procedure in 17 healthy bone marrow donors. Each sample was assayed for its content of red blood cells, nucleated cells, CD2+, CD4+, CD8+, CD19+, HLA-DR+, CD56+, CD13+, CD33+, CD34+ and KiM8+ cells and CFU-GM. On the basis of data obtained, we estimated that the first 1.0 ml samples had 8.0 +/- 5.2% (SD) and the transplant samples 20.8 +/- 8.5% contamination with nucleated blood cells. The calculation revealed that both types of bone marrow samples had 100% volume contamination with peripheral blood, i.e. that bone marrow cells were aspirated within blood fluid volume. Nucleated cell concentration was 3-fold, and CFU-GM concentration 10-fold lower in the transplant than in the first-puncture 1.0 ml bone marrow samples. Various marker-positive cells appeared in transplant samples in concentrations that depended on their abundance in the first-puncture 1.0 ml and blood samples. Taken together, our data suggest that bone marrow harvesting would be substantially improved if individual aspirates were small in volume and taken from bone puncture sites as distant as possible.     

Red blood cell phenotyping is a sensitive technique for monitoring chronic myeloid leukaemia patients after T-cell-depleted bone marrow transplantation and after donor leucocyte infusion

Fifteen consecutive patients with Philadelphia chromosome (Ph)-positive chronic myeloid leukaemia (CML) who relapsed from T-cell-depleted bone marrow transplantation (BMT) were successfully treated with donor leucocyte infusions (DLIs). Chimaerism was analysed using red blood cell phenotyping (RCP), and the results were compared with cytogenetic analysis and outcome of qualitative and quantitative polymerase chain reaction (PCR) for breakpoint molecules. In all patients, an increase in autologous erythrocytes and/or a decrease in donor red cells indicated relapse. Donor erythrocytes started to increase from 4 to 20 (median 12) weeks after DLI. At 6 and 12 months after DLI, complete donor chimaerism was found in 11 and 15 patients, respectively, and all patients were in cytogenic remission. A high percentage of autologous red cells at the time of DLI predicted pancytopenia. During relapse and after DLI, the percentage of autologous red cells was strongly correlated with the reappearance and disappearance of Ph-positive metaphases (r = 0.90; P < 0.001 and r = 0.96; P < 0.001 respectively). The same was true for the correlation between the percentage of autologous red cells and positivity/negativity in PCR for Bcr-Abl breakpoint molecules (r = 0.94; P < 0.001). A normalized Bcr-Abl dose of greater than 10-3 in real-time quantitative PCR correlated well with relapse and the presence of autologous red blood cells (r = 0.77; P < 0.001). We conclude that RCP is a sensitive, easy to perform and fast technique for the prediction of pending relapse after BMT for CML. RCP also predicts the response to DLI and the occurrence of bone marrow aplasia after DLI.     

Preparation of red-blood-cell-depleted marrow for ABO-incompatible marrow transplantation by density-gradient separation using the IBM 2991 blood cell processor

Thirty patients who had major ABO blood group incompatibility with their HLA-matched donors underwent allogeneic marrow transplantation after removal of red blood cells (RBC) from donor marrow by Ficoll-Diatrazoate (F-D) separation using the IBM 2991 blood cell processor. We selected the IBM 2991 because we were interested in obtaining information about RBC-depleted mononuclear cells for monoclonal antibody and complement incubation of marrow. The median residual marrow RBC volume was 2.6 ml (1.2% of the original volume) and marrow infusion was well tolerated in every instance. The median doses of nucleated and mononuclear cells were 8.7 X 10(7)/kg and 2.2 X 10(7)/kg recipient weight, respectively, representing median marrow total nucleated and mononuclear cell losses of 63.4% and 52.9%, respectively. The median CFU-GM loss was 52.4%. Four patients died 13-21 days after marrow infusion and were unevaluable for engraftment. One patient failed to achieve engraftment and received a second transplant on day 36 from a second donor. One patient with myelofibrosis had poor engraftment and died on day 177 with low peripheral blood counts. For evaluable patients, no significant differences were observed in the rate of recovery of peripheral blood granulocyte or platelet counts between those receiving RBC-depleted marrow or ABO-matched unprocessed marrow. However, posttransplant red cell transfusion requirements were increased and transfusion independence delayed in patients receiving RBC-depleted marrow as compared to patients receiving unprocessed marrow. We concluded that red cell depletion using the IBM 2991 was effective in removing red cell, but resulted in significant and variable hematopoietic cell losses which may have contributed to graft failure in at least one patient. Such cell losses appear to be inherent in both manual and semiautomated methods for F-D cell separation.     

Passive transfusion of human chorionic gonadotropin from plasma donated during pregnancy

Although fresh frozen plasma (FFP) prepared from autologous blood donated during pregnancy has frequently been given to homologous recipients at our institution, one transfusion resulted in an unanticipated diagnostic dilemma. A 31-year-old woman with disseminated intravascular coagulation of unclear etiology was transfused with multiple units of FFP, including 2 from pregnant autologous donors. A serum human chorionic gonadotropin (HCG) assay, performed because of the possibility that the patient's illness was a complication of unrecognized pregnancy, was positive using a blood sample drawn 7 h after the transfusions. An extensive evaluation was completed before the possibility of passive transfer of hormone from blood products was considered. Retrospective testing of serum samples established that HCG appeared in the patient's serum only after the first FFP transfusion from a pregnant autologous donor. In 8 other recipients of 1 or 2 units of FFP from pregnant autologous donors, post-transfusion HCG levels ranged between 96 and 1,750 mIU/ml. Of 15 recipients of packed red blood cells from pregnant autologous donors, only patients with renal failure or recipients of multiple units developed positive HCGs, which were always less than or equal to 85 mIU/ml. The differential diagnosis of a positive pregnancy test in a recently transfused individual should include the possibility of passively acquired hormone.     

Rapid and automated processing of bone marrow grafts without Ficoll density gradient for transplantation of cryopreserved autologous or ABO-incompatible allogeneic bone marrow.

The growing number of BMTs has increased interest in safe and standardized in vitro bone marrow processing techniques. We describe our experience with a rapid automated method for the isolation of mononuclear cells (MNC) from large volumes of bone marrow using a Fenwal CS-3000 cell separator without employing density gradient materials. Forty bone marrow harvests with a mean volume of 1650 +/- 307 ml were processed. A mean of 75 +/- 34% (50 percentile range 54-94%) of the original MNCs were recovered in a volume of 200 ml with only 4 +/- 2% of the starting red blood cells (RBC). Removal of granulocytes, immature myeloid precursors and platelets proved to be sufficient to permit safe cryopreservation and successful autologous BMT (n = 25). Allogeneic BMT (n = 14, including three major ABO-incompatible) could be performed without additional manipulation. In both groups of patients timely and stable engraftment comparable to historical controls receiving Ficoll gradient processed autologous (n = 17) or unprocessed allogeneic BMT (n = 54) was observed. Moreover, 70 +/- 14% of the RBC could be recovered from the grafts. They were used for autologous RBC support of donors, rendering unnecessary autologous blood pre-donations.