Between April 1985 and March 1987 130 children and adolescents up to 18 years of age with first relapse of acute lymphoblastic leukemia (ALL) were registered on the stratified and randomized multicentric trial ALL-REZ BFM 85 designed for patients pretreated with intensive front-line therapies. Stratification criteria were time and site of relapse: bone marrow (BM) relapse on or up to 6 months after stopping front-line therapy (group A), BM relapse beyond 6 months after therapy (group B), and isolated extramedullary relapse at any time (group C). Treatment consisted of alternating courses of polychemotherapy including randomly administered high- or intermediate-dose methotrexate (HDMTX:12 g/m2 as 4-hour infusion; IDMTX: 1 g/m2 as 36-hour infusion). During maintenance therapy the patients received daily oral thioguanine and biweekly intravenous (IV) MTX. The overall second complete remission (CR) rate was 92% (groups A, B, and C: 88%, 92%, and 100%), and the probability of event-free survival (EFS) at 6 years is 0.31 +/- 0.04 (groups A, B, and C: 0.18 +/- 0.05, 0.30 +/- 0.07, and 0.72 +/- 0.11). HDMTX did not prove to be superior to IDMTX, which led to premature stopping of randomization. Risk factor analyses showed early relapse, particularly BM relapse within 18 months, and T-cell phenotype to be independent predictors of poor outcome. The incidence of central nervous system (CNS) relapses following BM relapse was 19%, indicating that reprophylaxis to the CNS with IV/intrathecal (IT) MTX was insufficient. For 17 children who received bone marrow transplantation in second CR from HLA-compatible siblings the EFS was 0.53 +/- 0.12 at 5 years. Their outcome was not influenced by the above-mentioned risk factors. With the proposed treatment regimen long-lasting second remissions can be achieved in about one third of patients even after intensive front-line treatment.

One hundred ninety-three asymptomatic patients with hereditary coagulation disorders and human immunodeficiency virus (HIV) infection were studied in a controlled trial of zidovudine (ZDV) versus a placebo (with an average of 9.7 months on study). Pretreatment characteristics were well balanced between the placebo and drug-treated groups, including CD4 distributions, types of clotting disorders, transaminase abnormalities, and use of various hemostatic agents. At the time of analysis, 161 patients either were still receiving treatment or had previously reached an endpoint of disease progression while receiving treatment. Twenty-five patients withdrew voluntarily. The toxic effects noted included granulocytopenia and anemia, especially in older patients, and subjective symptoms of asthenia, malaise, and nausea, consistent with the known consequences of treatment with 300 mg ZDV five times daily. There was a trend toward more diagnoses of acquired immunodeficiency syndrome (AIDS), advanced or early AIDS-related complex (ARC), single ARC symptoms, or death in placebo recipients as compared with those receiving ZDV (22 v 13). Because older patients with hemophilia have more rapid disease progression, the same efficacy analysis was performed in the 89 patients aged more than 30 years who were receiving treatment. In this subgroup, there was a similar trend (11 v 6). With regard to the most advanced problems of the infection among the older patients, there were five patients who were newly diagnosed with AIDS or died in the placebo group versus none in the ZDV group (P = .02) among the older patients. The pretreatment distribution of CD4 counts for the placebo and ZDV groups were similar, but patients aged more than 30 years had significantly (P less than .049) fewer CD4 cells than patients aged less than 30 years. A beneficial ZDV effect is also supported by a trend toward higher CD4 counts (a 48-cell increase in the ZDV group at 24 weeks as compared with a four-cell increase in the placebo group) and a significant (P = .03) difference in weight gain in the ZDV patients aged more than 30 years (8 pounds) as compared with the older placebo patients (aged more than 30 years) (2 pounds) at week 24. The findings in the asymptomatic hemophilic patients aged more than 30 years support a useful effect of ZDV, which is similar to observations in the larger study of its use in asymptomatic, nonhemophilic patients.

One hundred twenty-seven patients with acute myelocytic leukemia (AML) were given busulfan 4 mg/kg on each of 4 days and cyclophosphamide 60 mg/kg on each of 2 days (BuCy2) followed by allogeneic bone marrow transplantation from an HLA-identical or one antigen disparate sibling. For 71 patients in first complete remission, 23 in second complete remission or initial relapse, and 33 patients with primary refractory disease, second or subsequent relapse, or a preceding hematologic disorder, the 3-year leukemia-free survival (LFS) is 63.1%, 32.6%, and 24.2% respectively. The actuarial probability of relapse for each group is 14.1%, 40.6%, and 61.0%. In multivariate analyses, relapse and decreased LFS were associated with advanced disease phase and with M4/M5 French-American-British classification. The LFS of first remission patients was adversely associated with a short time interval from diagnosis to transplantation. This study indicates that BuCy2 is an attractive preparative regimen for marrow transplantation in patients with AML and that prognostic factors for relapse and LFS are similar those described for regimens containing total body irradiation.

The pathogenic effect of cytomegalovirus (CMV) infection on the hematopoietic recovery after bone marrow transplantation (BMT) was retrospectively studied in 87 recipients of (nonpurged) autologous BMT and in 56 recipients of allogeneic BMT from HLA-identical siblings. Indications for autologous BMT were lymphomas or acute leukemias and for allogeneic BMT various malignancies or aplastic anemia. Patients were divided for the study in two groups, CMV-positive and CMV-negative on the basis of the CMV status pretransplant, and CMV-negative patients were kept CMV-negative by the local transfusion policy. In allogeneic BMT recipients, platelet recovery was significantly slower in CMV-positive patients than in CMV-negative patients (platelets greater than 50,000 cells/microL after 41 days v 27 days, P = .007). This difference held true when patients with acute graft-versus-host disease above grade I were excluded (platelets greater than 50,000 cells/microL after 42 days v 24 days, P = .01). In autologous BMT, the negative effect on platelet recovery was present in patients with lymphomas, but absent in patients with acute leukemias. Patients with acute leukemias had a very delayed recovery of platelets and granulocytes after autologous BMT, irrespective of the CMV status, probably due to the original stem cell disorder. Platelet recovery was significantly slower in CMV-positive autologous BMT recipients with lymphomas than in those not infected (platelets greater than 50,000 cells/microL after 36 days v 24 days, P = .0002). The presence of CMV infection had no effect on the recovery of granulocytes in autologous or allogeneic BMT. These data show that CMV infection causes delayed platelet recovery after BMT; however, in autologous BMT, the underlying disease (ie, acute leukemia) is more determinant for hematopoiesis after BMT.

Twenty-two children with hemophagocytic lymphohistiocytosis were treated with a chemotherapy regimen consisting of VP16-213, corticosteroids, and intrathecal methotrexate. A sustained clinical and biologic complete remission was obtained in 15 children and a partial remission in one child; six children died early of opportunistic infection (n = 4) or of disease progression (n = 2). Of the 16 children who were placed in first remission, 10 received maintenance chemotherapy alone, while six underwent bone marrow transplantation (HLA matched in five, HLA mismatched in one). Of the children who received chemotherapy alone, only two are in long-term remission after cessation of treatment. The remaining eight patients relapsed after a mean period of 5.4 months (range 2 to 8 months). Further treatment using the same regimen induced second remissions of short duration; death occurred after a median period of 2.3 months (range 0.5 to 6 months). A total of nine patients received allogeneic bone marrow transplantation (BMT). Among the six children transplanted in remission, four are in long-term unmaintained remission, 1 to 6 years after HLA-matched BMT. However, the relapse that occurred in one patient 1 year post BMT is difficult to interpret because the donor, the patient's 5-year-old sister, also developed the disease 1 year later. An HLA-nonidentical BMT resulted in unmaintained remission for 1 year, with autologous hematologic reconstitution followed by disease relapse. HLA-nonidentical BMT failed in three other patients with active disease at time of transplant. The poor long-term results of chemotherapy alone justify the use of related HLA-matched BMT in complete remission.

To explore the relationship between anti-interferon-alpha (anti-IFN-alpha) antibodies and loss of clinical responsiveness to IFN-alpha treatment, we examined sera from 59 patients with hairy cell leukemia who responded to therapy with recombinant IFN-alpha-2a (rIFN-alpha-2a). During the first 2 years of therapy, 10 patients developed rIFN-alpha-2a-neutralizing and 15 rIFN-alpha-2a-binding antibodies. Nine of the 59 initially responding patients became resistant to rIFN-alpha-2a and suffered a relapse of the disease at 7 to 24 months of treatment. All nine relapsing patients tested positive for both neutralizing and binding antibodies with titers above 400 INU/mL, while none of the antibody-negative patients relapsed. Six patients with detectable binding antibody titers below 400 INU/mL continued to respond to treatment. By measuring the IFN kinetics and the levels of the IFN-induced Mx-homologous protein in mononuclear cells after a single injection each of rIFN-alpha-2a and nIFN-alpha the IFN antibodies of eight of the nine resistant rIFN-alpha patients were found to be highly specific for rIFN-alpha-2a. Therefore, these eight patients were switched to natural IFN-alpha (nIFN-alpha) therapy at doses of 3 million IU, three times a week. All eight patients responded to treatment with nIFN-alpha, achieving durable objective responses similar to those obtained previously with rIFN-alpha-2a. These data clearly demonstrate that rIFN-alpha antibody-positive patients can effectively be treated with nIFN-alpha.

Verapamil was evaluated as a chemosensitizer for reversing multidrug resistance in multiple myeloma both in vitro and in clinical trials. Bone marrows from 59 myeloma patients in relapse were evaluated for several resistance parameters: expression of p-glycoprotein (MDR1), doxorubicin (Adriamycin) and vincristine sensitivity, and the ability of added verapamil to reduce resistance to the cytotoxic agents. We found that verapamil was capable of sensitizing myeloma cells that exhibited resistance to doxorubicin and vincristine in vitro, but did not enhance sensitivity of cells that were drug sensitive (P less than .001). Myeloma cells expressing MDR1 immunohistochemically tended to be more doxorubicin resistant in vitro than MDR1-negative cells. In the clinical trials, 22 patients with myeloma refractory to vincristine-Adriamycin-dexamethasone (VAD) were treated with VAD plus high-dose intravenous verapamil (Ve). Among the 22 patients treated with VAD/Ve, five achieved a partial remission (23%). The median relapse-free survival for the VAD/Ve responders was 5.4 months and their overall survival from the start of VAD/Ve was better than that of the nonresponders. Among the subset of 10 patients whose myeloma cells were MDR1 positive, four responded clinically (40%), whereas none of five patients with MDR1-negative myeloma cells achieved remission with VAD/Ve. We also observed that myeloma cells from three of four VAD/Ve clinical responders exhibited in vitro chemosensitization with verapamil, whereas in vitro verapamil chemosensitization was seen in only one of six clinical nonresponders. Our observations demonstrate that clinical reversal of multidrug resistance can be achieved in some patients with VAD-refractory myeloma with the use of verapamil. In addition to their value in drug development, in vitro tests of MDR1 expression and of chemosensitizers plus cytotoxic drugs on the patients' bone marrow myeloma cells may identify patients who will respond clinically to chemosensitizer-containing regimens. We anticipate that chemosensitizer regimens capable of inhibiting multidrug resistance will play an increasing role in the treatment of hematologic malignancies, including B-cell neoplasms such as multiple myeloma and the non-Hodgkin's lymphomas.

Seventy-seven cytomegalovirus (CMV)-seronegative marrow transplant patients were randomized in a prospective controlled trial comparing the use of leukocyte-depleted platelets plus CMV-seronegative red blood cells with standard unscreened blood products for the prevention of primary CMV infection during the first 100 days after transplant. Eligible patients included CMV-seronegative patients undergoing autologous transplant or seronegative patients undergoing allogeneic transplant for aplastic anemia or non-hematologic malignancy who had seronegative marrow donors. Patients and marrow donors were serologically screened for CMV and randomized before conditioning for transplant and followed for CMV infection with weekly cultures of throat, urine, and blood and with weekly CMV serologies until day 100 after transplant. Leukocyte-depleted platelets were prepared by centrifugation, a procedure that removed greater than 99% of leukocytes. There were no CMV infections observed in 35 evaluable treatment patients compared with seven infections in 30 evaluable control patients (P = .0013). There was no statistically significant difference in the mean number of platelet concentrates in the treatment patients (164 concentrates) compared with the control patients (126 concentrates). Leukocyte-depleted platelets plus CMV-seronegative red blood cells are highly effective in preventing primary CMV infection after marrow transplant.

Forty-seven patients with hematologic neoplasia received recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) by daily 2-hour infusion following allogeneic bone marrow transplantation from HLA-identical sibling donors in a phase I-II dose-escalation trial. Dose levels ranged from 30 to 500 micrograms/m2/d. At doses at or below 250 micrograms/m2/d, toxicity felt to be caused by rhGM-CSF was negligible. However, three of five patients treated with 500 micrograms/m2/d had unacceptable side effects caused by rhGM-CSF. Two different graft-versus-host disease (GVHD) prophylactic regimens were administered. Twenty-seven evaluable patients were administered regimens that did not contain methotrexate (MTX) (Group I) and reached an absolute neutrophil count of 1,000/microL by a median of day 14. In contrast, 18 patients who received GVHD prophylactic regimens containing MTX (Group II) reached an absolute neutrophil count of 1,000/microL on a median of day 20. Patients in Group I had fewer febrile days and, of those discharged, had shorter initial hospitalizations than patients in Group II. The overall incidence of severe acute GVHD (grade 2 or greater) in the rhGM-CSF-treated patients was 28% and was similar to that in historical "good risk" patients who did not receive rhGM-CSF. These preliminary data suggest rhGM-CSF is unlikely to exacerbate GVHD in HLA-identical sibling donor transplants and indicate the need for randomized trials of rhGM-CSF in allogeneic marrow transplant patients.

The efficacy, toxicity, and mechanism of effect of intravenous Anti-D (Winrho) were studied in 43 Rh+ patients with immune thrombocytopenia purpura (ITP) who had not undergone splenectomy and in three already splenectomized patients. The mean platelet increase for the 43 nonsplenectomized patients was 95,000/microL (median 43,000/microL). Children had greater acute platelet responses than did adults. Human immunodeficiency virus status and duration of thrombocytopenia did not affect response. Maintenance treatment was given to patients as needed: the average interval between infusions was 24 days. The three splenectomized patients had no platelet response whatsoever. Toxicity was minimal; infusions were completed in less than 5 minutes. The generally accepted mechanism of effect of Anti-D has been Fc receptor blockade by substitution of antibody-coated red blood cells for antibody-coated platelets. Evidence is presented suggesting that the effect of IV Anti-D is not limited to Fc receptor blockade, including: (1) no correlation of parameters of hemolysis with platelet increase; (2) a 48- to 72-hour delay before platelet increase; (3) a tendency of the change in monocyte Fc receptor I expression to correlate with platelet increase; and (4) increased in vitro production of antibodies to sheep red blood cells following IV Anti-D infusion.

We have used limiting dilution culture methods to determine the frequency of mitogen-responsive T cells in peripheral blood of patients after bone marrow autotransplantation, and have compared their responsiveness to that of allotransplant recipients and normal controls. Autotransplant patients were found to have low responder cell frequencies in tests for lymphokine-secreting helper function, and for IL-2 dependent proliferator and cytotoxic function. Multiple regression analysis showed that function was lower in autotransplant patients than in allorecipients, and lower in male patients for all three functional assays. Patients with clinically significant infection tended to have lower proliferative function in both transplant groups and lower cytotoxic function in the allotransplant population. Graft-versus-host disease was associated with lower T-cell function, but was present only in the allotransplant group; therefore, it cannot account for the even lower levels of function observed in the autotransplant population. Because we observe deficits in T-cell regeneration in autotransplant recipients that are even more severe than in allorecipients, we postulate that cellular immunodeficiency after bone marrow transplantation may reflect limitations in thymic-dependent repopulation rather than an effect of genetic disparity between host and donor (eg, clinical or subclinical graft-versus-host).

We report the results of a preclinical study comparing four different purging protocols using a promyelocytic human cell line HL-60 and myeloid leukemic progenitor cells (colony-forming unit-leukemic [CFU-L]) from acute myelogenous leukemia (AML) patients assayed in semisolid culture. We studied the antileukemic effect of (1) Single-cycle complement-mediated lysis by two different monoclonal antibodies (MoAbs) (M195 [CD33] and F23 [CD13] 40 micrograms/mL), reactive with distinct antigens found on early myeloid cells and monocytes, used alone and in combinations; (2) 4-Hydroperoxycyclophosphamide (4-HC) (80 mumol/L or 100 mumol/L) alone; or (3) combined with VP-16 (5 micrograms/mL) and (4) a cocktail of 1 through 3 as above (combined immunochemotherapy). More than 4 logs of HL-60 tumor cell elimination were observed after 1 hour of incubation with both MoAbs plus 4-HC + VP-16 while the single treatment (immunotherapy or chemotherapy) provided 1.5 and 3.5 logs of colony-forming inhibition, respectively. When the same protocols were tested on cryopreserved leukemic cells from eight patients with AML, we observed a mean value of CFU-L inhibition of 92.3% +/- 2.5% SD, 95.5% +/- 1.4% SD, and 99% +/- 0.8% SD after MoAbs and complement lysis, 4-HC, and 4-HC + VP-16 treatment, respectively. The combined treatment of MoAbs and 4-HC + VP-16 produced more than 3-log reduction of CFU-L colony formation. By comparison, the mean recovery of committed normal bone marrow progenitors after incubation with MoAbs and complement was 12% for CFU-granulocyte-macrophage (CFU-GM), 22.9% for burst-forming unit erythroid (BFU-E), and the recovery following 4-HC + VP-16 treatment was 4.4% for CFU-GM and 5.6% BFU-E. In subsequent experiments, highly purified CD34+ blast cells, enriched by positive selection, and stimulated in liquid culture by cytokines (interleukin-1 [IL-1], IL-3, and combination of both) or MO-conditioned medium (MoCM), demonstrated that immunochemotherapy spares hematopoietic colony-forming cells earlier than day 14 CFU-GM, in vitro.

4'-Demethoxydaunorubicin (idarubicin [IDR]) is a new anthracycline that differs from its parent compound by the deletion of a methoxy group at position 4 of the chromophore ring. This minor structural modification results in a more lipophilic compound with a unique metabolite that has a prolonged plasma half-life as well as in vitro and in vivo antileukemia activity. To determine its activity in acute myelogenous leukemia (AML), 130 consecutive adult patients between the ages of 16 and 60 with newly diagnosed disease were randomized in a single institution study to receive either IDR in combination with cytosine arabinoside (Ara-C) or standard therapy with daunorubicin (DNR) and Ara-C. The trial was analyzed using the O'Brien-Fleming multiple testing design that allowed for periodic inspection of the data at specific patient accession points. After accrual of 60 patients per arm, analysis showed that patients who received IDR/Ara-C had a superior response compared with those who received standard therapy: 48 of 60 patients (80%) achieved complete remission on the former arm compared with 35 of 60 patients on the latter (58%, P = .005). Logistic regression analysis of factors associated with complete response indicated that treatment with IDR/Ara-C offered a significant advantage to patients who presented with a high initial white blood cell count compared with treatment with DNR/Ara-C. The degree of marrow aplasia was approximately the same on each arm as was nonhematologic toxicity. Overall survival for patients on the IDR/Ara-C arm was 19.5 months compared with 13.5 months on the DNR/Ara-C arm (P = .025) at a median follow-up of 2.5 years. We conclude that IDR/Ara-C can effectively replace standard therapy with DNR/Ara-C in adult patients less than age 60 with newly diagnosed AML.

A randomized trial was performed to compare two regimens of total body irradiation in patients with chronic myeloid leukemia treated by allogeneic marrow transplantation while in the chronic phase. All patients received cyclophosphamide 120 mg/kg followed by total body irradiation and marrow from HLA-identical siblings. Cyclosporine and methotrexate were used for prophylaxis against acute graft-versus-host disease. Fifty-seven patients were randomized to receive 2.0 Gy fractions of irradiation daily for 6 days and 59 were randomized to receive 2.25 Gy fractions daily for 7 days. The probabilities of relapse at 4 years were 0.25 for the 12.0 Gy group and 0.00 for the 15.75 Gy group (P = .008). The actuarial probabilities of survival and relapse-free survival at 4 years were 0.60 and 0.58 among the patients who received 12.0 Gy compared with 0.66 and 0.66 for those who received 15.75 Gy. The 4-year probabilities of transplant-related mortality were 0.24 and 0.34 respectively (P = .13) while the probability of moderate to severe acute graft-versus-host disease was 0.33 for the 12.0 Gy group and 0.44 for the 15.75 Gy group (P = .15). The lower relapse probability in the patients receiving the higher dose of total body irradiation did not result in improved survival because mortality from causes other than relapse was increased.

We report the safety and efficacy of 34 consecutive autologous bone marrow transplant (ABMT) procedures performed in adult patients with high-grade lymphoid malignancy after remission induction therapy. Fifteen patients with acute lymphoblastic leukemia (ALL) and six with high-grade non-Hodgkin's lymphoma (NHL) received pretransplant conditioning with intravenous (IV) melphalan and fractionated total body irradiation (TBI). Thirteen other patients with NHL were conditioned with melphalan alone, having previously received local involved field radiotherapy. Unmanipulated noncryopreserved autologous marrow was reinfused within 48 hours of harvesting. Engraftment occurred in all patients with medians of 10 days of neutropenia (neutrophils less than 0.5 x 10(9)/L), 4-day platelet transfusion requirement, 3 U packed RBC transfusion, and 18 days in hospital posttransplant. There were no procedure-related deaths. Actuarial disease-free survival in the 13 patients with ALL receiving autotransplant early in first remission is 48% with a median follow-up of 3 years. Two other ALL patients who had autotransplants after a period of maintenance therapy also remain in complete remission (CR). These results compare favorably with our 34% disease-free survival (DFS) in 15 allogeneic ALL transplant patients and 21% DFS in 19 patients on standard maintenance after a common induction schedule. No relapses have occurred in the 17 NHL patients transplanted in remission (median follow-up 2 years), but the two NHL patients who developed recurrent disease before ABMT died of progressive disease after temporary responses. We conclude that this method of ABMT results in rapid reengraftment with lack of toxicity and that the conditioning treatment used shows good efficacy against disease. It is applicable in high-grade lymphoid malignancy in first remission, and our results call into question the need for marrow purging in ALL and NHL patients transplanted in first remission.

Fifty patients with intermediate- or high-grade non-Hodgkin's lymphoma (NHL) who had relapsed after a complete remission induced by an Adriamycin-containing chemotherapy regimen participated in this prospective pilot study. The patients ranged in age from 16 to 60 years (median 42 years). All patients received dexamethasone, high-dose cytarabine, and cisplatin (DHAP) for two courses at 3- to 4-week intervals. Patients achieving a partial or complete response were scheduled to receive involved-field radiotherapy and high-dose carmustine, etoposide, cytarabine, and cyclophosphamide (BEAC), followed by autologous bone marrow transplantation (ABMT). Among 48 evaluable patients (ie, 1 was lost to follow-up and 1 had no measurable disease) 7 patients obtained a complete response (CR) and another 21 patients achieved partial response (PR), whereas the remaining 20 patients failed. One responder died of treatment-related toxicity, and six others declined ABMT. The patient with no measurable disease did not progress on DHAP and was submitted to ABMT. Twenty-two patients underwent ABMT [20 with BEAC and 2 with cyclophosphamide plus total body irradiation (TBI)] of whom 2 (9%) died of toxicity and 10 relapsed. One patient was a suicide at 28 months post-ABMT in CCR and 9 are alive disease-free 24 months to 32 months (median 30 months) post-ABMT. The actuarial 2-year event-free survival for patients undergoing transplantation is 40%. This prospective multicenter trial documents the ability of DHAP followed by ABMT to produce durable complete remission in a significant proportion of patients with relapsed aggressive NHL. Forty-four percent of all patients with relapsed lymphoma who entered the study actually underwent ABMT and 20% of the total group are projected to be long-term disease-free survivors.

Eighty-one patients with acute myeloid leukemia (ANLL, n = 44) or acute lymphoblastic leukemia (ALL, n = 37), aged 10 to 50 years were randomized to receive 1 mg/kg per day (n = 41, group A) or 5 mg/kg per day (n = 40, group B) of cyclosporine A (CyA) from day -1 to day +20 after bone marrow transplant (BMT). All patients received CyA orally thereafter. All patients were prepared with cyclophosphamide (CY) 120 mg/kg and fractionated total body irradiation (TBI), and received unfractionated BM from an HLA-identical sibling. The two groups were comparable for diagnosis, disease status, French-American-British (FAB) classification, WBC count at diagnosis, cytogenetic abnormalities, extramedullary disease before BMT, donor/recipient age and sex, number of cells infused, and number of days with intravenous (IV) CyA. Median follow-up for surviving patients in group A was 983 v 632 days in group B. Patients in group A had lower serum levels of CyA (295 v 686 ng/mL, P = .004), lower bilirubin levels (1.9 v 2.6 mg/dL, P = .07), lower creatinine levels (0.9 v 1.4 mg/dL, P = .06), and a lower proportion of CD8+ cells in the peripheral blood (PB) within day +21 (19% v 28%, P = .07). First day to 0.5 x 10(9)/L neutrophils was comparable in the two groups (13 v 14 days; P = .1). In a Cox model, the actuarial risk of acute graft-v-host disease (GVHD) grade II+, after stratification for age (less than 20 years greater than) was significantly lower in group B patients (0.54, P = .04). The actuarial risk of developing chronic GVHD was comparable (P = .9). Actuarial transplant-related mortality (TRM) at 240 days was 28% and 26% (P = .8) in group A and B: the major cause of death was GVHD in group A (P = .02) and multiorgan toxicity in group B (P = .07). The actuarial risk of relapse at 2 years overall was 20% in group A and 52% in group B (P = .001); it was 9% v 43%, respectively, for patients in first remission (P = .0001) and 48% v 63% for patients in non-first complete remission (CR) (P = .1). Actuarial 2-year disease-free survival (DFS) in group A and B was 58% v 32% (P = .02) for all patients, 71% v 35% (P = .01), in first remissions, and 30% v 23% (P = .2) in advanced disease.(ABSTRACT TRUNCATED AT 400 WORDS)

In a double-blind study, we compared the postoperative (post-op) blood loss in 161 children undergoing open heart surgery with cardiopulmonary bypass whose immediate post-op transfusion requirements were met with either very fresh whole blood (VFWB), 24- to 48-hour-old whole blood or reconstituted whole blood (packed red blood cells, fresh frozen plasma [FFP], and platelets). Assignment to treatment groups was not strictly random but dependent, in part, on the ability of families to provide directed donors for fresh blood. The three patient groups were comparable with respect to patient age, pre-op coagulation profiles (bleeding time, prothrombin time, activated partial thromboplastin time, platelet count, fibrin split products, fibrinogen, and platelet aggregation tests) difficulty of operative procedures and time spent on CPB. Mean 24-hour post-op blood loss in milliliters per kilogram was 50.9 +/- 9.3 in the VFWB group, 44.8 +/- 6.0 in the 24- to 48-hour-old group, and 74.2 +/- 8.9 in the reconstituted group (p = .03). When blood loss was compared in the 93 children less than 2 years of age, mean blood loss was 52.3 +/- 10.8 in the VFWB group, 51.7 +/- 7.4 in the 24- to 48-hour-old group, and 96.2 +/- 10.7 in the reconstituted group (P = .001). For subjects who had received reconstituted blood, 30-minute and 3-hour post-op platelet aggregation responses to adenosine diphosphate (10 mumol/L) and 30-minute aggregation response to epinephrine (2.5 mumol/L) were more depressed than in the VFWB and 24- to 48-hour groups (P less than .001, P = .005, and P = .02). Comparison of other post-op coagulation tests could not explain the increased blood loss in the reconstituted group. We conclude that the transfusion of less than 48 hours old whole blood is associated with significantly less post-op blood loss than the transfusion of packed red blood cells, FFP, and platelets in children under 2 years old who underwent complex cardiac surgery. The blood losses associated with the transfusion of VFWB and 24- to 48-hour-old blood are comparable and may be, in part, due to better functioning platelets.